JPH0694899B2 - Automatic transmission control device - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an automatic transmission, and more specifically, to expert operation performed in a conventional manual transmission by applying fuzzy control theory. Judgment
The present invention relates to a control device for an automatic transmission that enables control similar to operation.

(Prior Art) Conventionally, a manual transmission is used for a vehicle transmission,
The driver had to shift gears by operating the clutch pedal and the shift lever while judging the gear shift timing according to the driving condition while considering the surrounding conditions. However, since such manual shifting is complicated, an automatic transmission has been developed, and in recent years, it has been installed in a majority of passenger cars sold. Thus, in such a control device for an automatic transmission, a shift valve is provided in the hydraulic circuit, a throttle pressure proportional to the throttle opening is applied to one end of the valve, and a governor pressure proportional to the vehicle speed is applied to the other end. Is actuated to supply / disconnect the hydraulic pressure to the gear clutch according to the pressure ratio between the two, and the gears are automatically switched. Also, with the subsequent electronic control, a control device was constructed with a microcomputer,
The shift map stored in the memory is searched from the throttle opening and the vehicle speed to detect the shift point, and the solenoid valve is excited / de-excited to drive the shift valve to switch gears.

Therefore, in the conventional automatic control device, if the previous manual transmission was used, the shift time point that the driver himself / herself judged and operated is uniquely determined from the throttle opening and the vehicle speed.
It was undeniable that an unnatural shift would occur.
For example, when the driver returns the throttle opening to the cruise opening in the same way as when driving on level ground when climbing, the vehicle may shift up depending on the traveling vehicle speed, so the marginal driving force is insufficient and the accelerator pedal is depressed again. As a result of downshifting, repeated downshifts and upshifts occur, which gives the driver a busy feeling. Such an inconvenience occurs when the camper or the like is towed, when the vehicle weight increases due to loading or the like, or when traveling at high altitude where the engine charging efficiency deteriorates.

Considering why the driver depresses the accelerator pedal to open the throttle valve, we expect that the vehicle will follow the driver's acceleration request indicated by opening the throttle valve. Nothing else. That is, the above-mentioned inconvenience is caused in other words because the shift command is issued by the control device although the surplus driving force is reduced and the controllability of the vehicle is not sufficiently ensured. Therefore, for that purpose, it is necessary to surely grasp the driving force and the running resistance in the control device, and to confirm that the driving force exceeds the running resistance and that there is a surplus driving force, and to shift up. Because it is not done like.

From this point, a technique described in Japanese Patent Laid-Open No. 60-143133 has recently been proposed. In this technique, the torque required by the driver is obtained from the accelerator pedal depression amount, and the separately calculated climbing resistance is subtracted. To calculate the required acceleration. Furthermore, a shift map corresponding to the climbing resistance detected from the plurality of best fuel consumption shift diagrams is selected, and the throttle opening is set to realize the required acceleration from the constant acceleration traveling trajectory data on the shift diagram. It is configured to maintain the acceleration before the change by controlling and further searching the shift line diagram from the changed throttle opening and the changed vehicle speed to determine the shift.

(Problems to be Solved by the Invention) However, in the above-mentioned conventional technique, although the shift determination is performed in consideration of the torque required by the driver, the shift determination is based on the preset shift diagram. It is done based on the fact that it can only respond to the set situation, and in any case, the shift time point is uniquely determined from the throttle opening and the vehicle speed. It was hard to avoid criticism similar to technology.

If this is a manual transmission vehicle, the driver should recognize that he / she is climbing when he / she performs a gear shifting operation, and avoid inadvertent upshifting. That is, in a manual transmission vehicle, the driver grasps the driving state of the vehicle including the situation of four circles,
The shift time should be determined by recognizing the driving force output from the vehicle and also predicting the increase or decrease in the driving force when shifting, and selecting various learned rules of thumb. That is,
The above-mentioned inconvenience is due to human judgment /
This is because the motion is disregarded and is not reflected during control. That is, the conventional automatic shift control technique basically mechanically determines the shift time point from the throttle opening and the vehicle speed, and does not judge the shift time point by grasping the operating state of the vehicle with multiple variables. Therefore, it is unavoidable that the above-mentioned inconvenience occurs.

Therefore, an object of the present invention is to eliminate the above-mentioned drawbacks in the prior art, and to incorporate the shift operation that the expert driver has judged and operated in the manual transmission vehicle into the automatic shift control by applying the fuzzy control theory. Therefore, it is an object of the present invention to provide a control device for an automatic transmission that enables a determination / operation similar to human decision making.

More specifically, it is an object of the present invention to provide a control device for an automatic transmission, which is capable of determining a control value using a fuzzy control theory and appropriately determining the control value.

(Means and Actions for Solving the Problems) In order to achieve the above object, the control device for an automatic transmission according to the present invention is, as shown in FIG. 1, an engine speed, a throttle opening, a throttle opening speed, a vehicle speed. A driving state detecting means 1 for detecting a driving state of an engine or a vehicle including a vehicle acceleration; and a change in a driving state which may occur if a shift is made to the gear stage for all gear stages capable of shifting from the current stage, Operating state predicting means 2 for predicting a change in the surplus driving force and a change in the engine speed obtained from the running resistance calculated based on the road surface inclination angle, the operating state to be detected and the operating state to be predicted. Membership function setting means 3 for setting a membership function in the fuzzy set in advance for changes, a plurality of the detected values and predicted values and the set membership function. A gear shift stage determining means 4 for selecting one of the fuzzy production rules to determine the shift gear stage, and a driving means 5 for driving the transmission mechanism according to the output of the shift gear stage determining means. At the same time, the shift gear stage determining means is configured to prohibit selection of a specific fuzzy production rule in a predetermined operating state.

Embodiment An embodiment of the present invention will be described below with reference to the accompanying drawings.

FIG. 2 is a schematic view showing an overall control device for an automatic transmission according to the present invention. When the description is made with reference to FIG. An intake passage 12 is connected to the engine body 10, and an air cleaner 14 is attached to the tip side of the intake passage 12. Thus, the intake air introduced from the air cleaner 14 is
The flow rate is adjusted through a throttle valve 16 that operates in conjunction with an accelerator pedal (not shown) on the floor of the vehicle driver's seat to reach the engine body. A fuel injection valve (not shown) is provided at an appropriate position near the combustion chamber (not shown) of the intake passage 12 to supply fuel, and intake air is mixed with the fuel and enters the combustion chamber. After being compressed by (not shown), it is ignited through a spark plug (not shown) to explode and drives a piston. The piston driving force is converted into rotational movement and taken out from the engine output shaft 18.

A transmission 20 is connected to a rear stage of the engine body 10, and an engine output shaft 18 is connected to a torque converter 22 there, and is connected to a pump impeller 22a thereof. The turbine runner 22b of the torque converter 12 is connected to a main shaft (mission input shaft) 24. Main shaft
A counter shaft (mission output shaft) 26 is juxtaposed to 24, and a first speed gear G1 and a second speed gear G are provided between both shafts.
A second- and third-speed gear G3, a fourth-speed gear G4, and a reverse gear G _R are provided, and a multi-plate hydraulic clutch C _L1 , C _L2 , C _L3 , C _L4 is provided for each gear (the reverse gear clutch is not shown). Are omitted for simplification).
A one-way clutch 28 is attached to the first speed gear G1. An oil passage 30 connecting a hydraulic source (not shown) and a tank (not shown) is connected to these hydraulic clutches, and A and B two shift valves 32 and 34 are inserted in the middle of the oil passage 30. The shift valve changes its position depending on the energized / de-energized state of the two electromagnetic solenoids 36, 38 to control the supply / discharge of the pressure oil to the clutch group. The torque converter 22 includes a lockup mechanism 40, and directly connects the turbine runner 22b and the engine output shaft 18 in accordance with a command from a control unit described later. Thus, the counter shaft 26 is connected to the rear axle 44 via the differential device 42, and the rear wheels 46 are attached to both ends thereof. The engine body 10, the transmission 20, and the differential device 42 are mounted on a chassis (not shown), and a frame (not shown) is mounted on the chassis to form a vehicle.

Thus, a throttle sensor 50 including a potentiometer for detecting the opening degree is provided near the throttle valve 16 of the intake passage 12, and a rotating portion such as a distributor (not shown) near the engine body 10 is provided. A crank angle sensor 52 composed of an electromagnetic pickup or the like is provided at the position to detect the crank angle position of the piston and output a signal for each predetermined crank angle. Further, an inclination angle sensor 54 for detecting an inclination angle of the vehicle, that is, an inclination angle of a road surface on which the vehicle travels is provided at an appropriate position of a body frame (not shown), and a reed switch is provided at an appropriate position of the transmission 20. A vehicle speed sensor 56 including the above is provided to detect the traveling speed of the vehicle. The outputs of these sensors 50, 52, 54, 56 are sent to the shift control unit 60. Further, the outputs of the range selector switch 62 for detecting the selected position of the range selector and the shift position switch 64 for detecting the shift position (gear stage) are also sent to the control unit.

FIG. 3 is a block diagram showing the details of the shift control unit 60. As shown in FIG. 3, the outputs of the throttle sensor 50 and the tilt angle sensor 54 are first input to the control unit and then to the level conversion circuit 68. And amplified to the appropriate level,
Input to the microcomputer 70. The microcomputer 70 includes an input port 70a, an A / D conversion circuit 70b, a CPU
70c, ROM70d and RAM70e and output port 70f and a group of registers and counters (both not shown),
The output of the level conversion circuit 68 is input to the A / D conversion circuit 70b, converted into a digital value, and temporarily stored in the RAM 70e. Similarly, the outputs of the crank angle sensor 52 and the like are also waveform-shaped by the waveform shaping circuit 72 in the shift control unit,
It is input into the microcomputer via the input port 70a and temporarily stored in the RAM 70e. The CPU 70c determines a shift command value based on these measured values and the calculated value calculated from them, and sends it to the first output circuit 74 and / or the second output circuit 76 from the output port 79f as described later. The electromagnetic solenoids 36 and 38 are excited / de-excited to switch or hold the shift position. It should be noted that the gear shift is performed so that, for example, when both solenoids are de-energized (OFF), the fourth speed gear is engaged. However, the shift operation itself via such an electromagnetic solenoid is known. However, since it is not a feature of the present application, detailed description thereof will be omitted.

Next, the operation of the present control device will be described with reference to the flow charts of FIG.

Here, the features of the present control device will be briefly described before starting a specific description. The feature of the control device according to the present invention is that the fuzzy control theory is applied to shift timing in a manner close to human decision making. The point is that it is configured to determine. That is, the feature of the present invention is not the configuration of the control device itself, but the operation of the control device, that is, the control method. Since the fuzzy control theory itself is being applied to various fields in recent years, its detailed description will be omitted, but simply speaking, the state recognition of the controlled object is vaguely understood, and based on the state recognition. The control rules that determine control values (called "production rules") themselves are also linguistically expressed in the form of "if-if", and in the production rules, the criteria for judging situations or the contents of operations. Is treated as an ambiguous amount (shown as a fuzzy label or variable) and is quantified by the membership function. That is, while using fuzzy information performed by humans, a flexible and highly adaptive control operation is modeled by fuzzy theory, and a control value is calculated by using fuzzy reasoning. Since the knowledge possessed by human beings can be easily expressed, it is easily adapted to a so-called expert system in which the knowledge and judgment of a skilled person are incorporated into a computer system. This controller is based on such a theory.

Therefore, in this control device, when designing the control system of the automatic transmission, work such as creation of production rules necessary for introducing fuzzy control theory is performed, and at the time of actual running, the control value is calculated based on the control algorithm. It is determined and is specifically performed as follows.

(1) Creation of production rules As will be described later, an appropriate number of rules expressed in language, such as "If you are aiming for a cruise at a reasonable speed, do not shift" are created. When creating this rule, the expert driver's judgment and operation in the manual transmission vehicle are analyzed, and then heuristic rules that are inductive are selected.

(2) Determination of parameters and membership functions At the same time, it is determined from what parameters the state of the controlled object is recognized, and the parameters (variables) used for each of the production rules are selected. The evaluation function is determined by defining the membership function of the parameter (the state expressed by such membership function is called fuzzy label). As the parameter, the control device uses an actual measurement value which is a physical quantity detected through a sensor and a calculated value obtained by differentiating the actual measurement value. Specifically, the engine speed,
Using the throttle opening, vehicle speed, throttle change amount, acceleration, etc., as shown in FIG. 10, an appropriate waveform (membership function) with parameters is given to the horizontal axis (hereinafter referred to as “domain”) in the coordinates. , Values from "0" to "1.0" on the vertical axis (referred to as "membership value (grade)")
Attach.

The above is the preparatory work for vehicle design. Incidentally, in the preparation stage, along with this, selection of a sensor for detecting the determined parameter, storage of a control rule in the memory of the microcomputer of the control unit, storage of an instruction of a calculation procedure, and the like are performed.

(3) Control during actual running CPU7 in the microcomputer while running
In 0c, a parameter is detected (calculated), a control rule is referred to, fuzzy reasoning is performed to select one of the control rules, and a control result, for example, the first speed up is determined based on the selected control rule, and then a predetermined electromagnetic solenoid 36 , 38 are energized / de-energized to engage the gear in the first speed. In this fuzzy reasoning, the membership value is calculated for the parameter related to each control rule, and the minimum value is used as the evaluation value of that control rule, and the control rule with the largest evaluation value among all control rules. Select. Such mini-max operation itself is often used in fuzzy inference.

Next, the operation of the control device will be described with reference to FIG. FIG. 4 is a main routine flow chart schematically showing the operation of this control device. It should be noted that this program is started at an appropriate cycle such as 10 ms to 40 ms.

First, in S10, the actual measurement value detected by the sensor at the time of starting the program this time is read and stored. As the actual measured values, the engine speed Ne (rpm) (calculated by integrating the output of the crank angle sensor 52 for a predetermined time), the vehicle speed V (km / h), the throttle opening θ _TH (degree), the traveling Road surface inclination angle tan θ, current shift position (gear stage) signal S _FT0 (calculated from the ratio of the input shaft speed to the output shaft speed of the mission, or the engine speed, throttle opening, vehicle speed, etc.),
Elapsed time after shift t _SFT (s) (This is not the sensor output, but is obtained by measuring the time with the timer counter of the microcomputer. That is, in the microcomputer, when the shift command is issued, the appropriate bit of the flag register is set. Since it is turned on, the elapsed time since it was turned on is measured) and the range position signal P _RANGE is used.

Subsequently, after confirming that the range selector is in the D range in S12, it is determined in S14 whether or not the gear shifting operation is currently being performed. This determination work is performed with reference to the shift command flag described above. If it is confirmed in S14 that the gear shift is not in progress, the flow proceeds to S16 to determine the gear shift command value. This will be described later. In addition, when negative and affirmative in S12, 14, this program ends immediately.

FIG. 5 shows a subroutine flow for determining the gear shift command value.
It is a chart. To explain according to the figure, first, in S100, the vehicle speed V (km / h) and the throttle opening θ _TH (degree) are read out from the sensor output values detected at the time of the previous program startup, and the acceleration α (km / h / s) ( Vehicle speed deviation) and throttle change amount Δθ _TH (degree / s) and acceleration change amount Δα (km
/ h / s / s) is calculated. That is, as shown in FIG. 6, the acceleration and the throttle change amount are the values at the time of starting the program this time (time n) and the values at the time of the previous program start (time n-
(1) and the deviation (first derivative value divided by unit time n- (n-1)). The acceleration change amount is obtained by further differentiating the acceleration with respect to time. In the actual calculation, the acceleration is “km / h / 0.1s” and the throttle change is “degree”.
/0.1s "and the acceleration change amount is calculated as" km / h / 0.1s / 0.1s ".

Then, in S102, the post-shift rotation speed Ne-SFT is calculated.
Since this is a gear other than the current gear, that is, four forward gears in the present embodiment, it is presumed how the engine speed changes when it is assumed that the gear is shifted from one gear to another three gears. This is calculated according to the subroutine flow chart shown in FIG.

Referring to FIG. 7, first, the vehicle speed V is read in S200, then the current shift position S _FT0 is read in S202, and S204
Initialize the counter value S _FT1 that counts the shift position after shifting with S _FT1 = 1 (S _FT1 = 1 means the first speed). Next, in S206, the first speed and the current shift position
The maximum number of gears that can be downshifted is calculated by _obtaining the difference from _SFT0 . As shown in FIG. 8, assuming that the vehicle is currently in the third speed, this is the second speed. Subsequently, in S208, the gear ratio G / R of the first speed is read from the data stored in advance in the ROM 70d of the microcomputer, and it is confirmed in S210 that the first speed is not the current shift position. The post-shift rotation speed Ne-SFT is calculated as follows.

Ne-SFT = (V × 1000 × G / R) / (60 × 6.28 × r) [rp
m] ∵V: vehicle speed, 1000,60,6.28: constant for calculating vehicle speed from gear ratio, G / R: gear ratio (for the gear), r: tire effective radius (m) Then, in S214 The number of gears is calculated from the difference between the gear (first gear) and the current gear, and the work is repeated until the fourth gear is reached (S216, 218), and after reaching the highest gear, it is possible to go up in S220. The number of shift stages is calculated and the process ends. FIG. 8 shows a calculation example of these.

Returning to FIG. 5 again, the fuzzy production rule is subsequently searched in S104. Figure 9 shows the subroutine
Although it is a flow chart, the rules used in this embodiment will be briefly described with reference to FIG. 10 before the description of FIG. As described above, the rules, the used parameters, the membership function, and the like are determined in advance when the control system of the vehicle is designed.

FIG. 10 is an explanatory diagram showing the fuzzy production rule in detail, but in the case of this embodiment, 44 rules are used. This will be described below.

Rule 1 Use fuzzy label Conclusion Throttle change Δθ _TH [without shift / 0.1s. Specifically, WOT / 8 / 0.1s. same as below.

WOT = 84 degrees] Acceleration α [km / h / 0.1s. Same below) Engine speed Ne [rpm. Same below) Rule Implication "If you are aiming for a cruise at a reasonable speed, do not shift." That is, if you are aiming for a cruise at a reasonable speed, there will be no change in throttle opening and acceleration. It should be constant, and if the engine speed is 2000 rpm or less, it is appropriate as a cruise.
Is considered to be appropriate as a cruise depending on the degree, and it is expected that it is not appropriate as a cruise if it is 4000 rpm or more.) Therefore, evaluate the degree of satisfaction from the actual driving condition, and use all other rules. This rule will be adopted when the evaluation exceeds satisfaction.

The evaluation method of such a rule will be described later with reference to FIG. 9, but briefly, it will be briefly described by searching for a position where the value of the parameter detected (or calculated) in rule 1 matches on the domain. And read the membership value (more specifically referred to as "grade") on the vertical axis that intersects the waveform. For example, the actual value is Δθ
TH = 0 (degree / 0.1s), α = 0 (km / h / 0.1s), Ne = 3000
(Rpm). Applying this to the fuzzy label in Fig. 10, the grade of ΔθTH is 1.0, as shown by the broken line.
The α grade is 1.0 and the Ne grade is 0.5. In such a case, the grade of the minimum value is an evaluation indicating the entire related fuzzy label in the sense that it satisfies all the other fuzzy labels at least in its range, and as a result, the evaluation of rule 1 is 0.5. Becomes Subsequently, the same evaluation is performed for all the rules, the rule having the highest evaluation value is adopted in the sense of having the highest degree of satisfaction, and the shift position is determined according to the conclusion indicated by the rule. Therefore, if rule 1 is adopted, it will not shift.

Next, rules 2 and onward will be described.

Rule 2 Fuzzy label Conclusion Engine speed Ne only 1st speed up Implication of the rule "In high rotation speed 1st speed up to protect engine" This is a rule of engine protection, engine speed 6000rpm
When entering the red zone that exceeds the limit, the first speed is increased to protect the engine. The expression “up 1st gear” means upshifting for 1st gear, for example, shifting to 3rd gear if it is 2nd gear, and does not mean shifting to 1st gear.
This also applies to other rules.

Rule 3 Fuzzy label Conclusion Elapsed time after shift tSFT [s] No shift, throttle change amount ΔθTH Implication of rule “Do not change immediately after shift” Driver does not push the throttle valve immediately after shift It is estimated that you do not have a shift intention,
For example, a dead zone of about 1.6 to 2.5 seconds is provided.

Rule 4 (Rules 5 and 6) Fuzzy label Conclusion Current shift position S _FT0 3rd down _Vehicle speed V [km / h. The same applies below. (2nd speed, 1st speed down) Throttle opening θTH [degrees].

same as below. Note that WOT = 84 degrees] Rule implication "To 1st speed when fully closed and extremely low vehicle speed" From Rules 4 to 6, shift down to 1st speed when the throttle is fully closed and the vehicle speed is extremely low. This is a rule when the present shift position is the fourth speed, the next rule 5 is the third speed, and the rule 6 is the second speed. These rules are the shift initial actions.

Rule 7 (Rule 8) Fuzzy label Conclusion Current shift position S _FT0 2nd speed down _Vehicle speed V (1st speed down) Throttle opening θTH Implication of rule "If fully closed / low vehicle speed, go to 2nd speed" This rule and the next The rule is similar to the rules 4 to 6, and an example in which the vehicle speed is shifted to the second speed when the vehicle speed is still low, even if the vehicle speed is not so low.

Rule 9 Fuzzy label Conclusion Throttle opening θTH Not shifted Throttle change amount ΔθTH Rule implication “If throttle returns abruptly, shift does not occur without intention to shift.” This is to prevent the control from relaxing. There are mainly control techniques.

Rule 10 Fuzzy label Conclusion Current shift position S _FT0 Up from first speed to second speed Acceleration change amount Δα [km / h / 0.1s / 0.1s. The same applies hereafter] Control toughness CT (2nd speed) Rotational speed after shifting Ne-SFT [rpm. The same applies hereafter] (Second speed) Acceleration α Throttle change amount ΔθTH Rule implication “Upshifting from first speed to second speed during acceleration is performed if acceleration is saturated and controllability after shifting is good.” If it is increasing and becoming saturated, and the throttle is not returned (while stepping on), the second speed is increased. In this case, the control toughness at the second speed, which is the destination stage, is searched from the fact that the speed is increased, and the post-shift rotation speed is also evaluated by referring to the calculated value at the second speed.

In addition, in this rule, the concept of control toughness is used as a parameter, and the situation in the case of upshifting is also predicted and then determined. Here, the control toughness will be described, and this is a word related to the coined words of the inventors, and is used to mean a "coefficient representing the appropriateness of the reaction of the vehicle to the change in the throttle opening θTH". This concept was devised from the point that the present invention aims to eliminate the busy feeling in which shifts are frequently repeated when climbing a plate or when towing a camper as described above. That is, the above-mentioned inconvenience is caused by the fact that the surplus driving force obtained by subtracting the running resistance, which is an external load of the vehicle, from the driving force is not sufficiently secured, and the reduction of the surplus driving force is caused by the driving force itself. Becomes noticeable at the time of shift up, where

This point will be explained with reference to the drive force diagram of FIG. 11. If the engine speed is currently running at Ne0, the difference between the fully open drive force and the amount of spare horsepower is as shown in the figure. Shown. In this case, the running resistance is higher than that when running on a flat road due to the addition of gradient resistance when climbing. Thus, when the throttle opening is returned to the cruise opening in this state, the conventional control apparatus automatically shifts up because the shift point is uniquely determined from the vehicle speed and the throttle opening. Therefore, the engine speed drops to Ne1, and the full-open drive force also drops, so the equivalent horsepower also decreases.
As a result, the downshift is performed again. As described above, the up, down, and up are repeated to give the driver a troublesome feeling, as described above.

That is, in this case, the running resistance is large with respect to the amount equivalent to the surplus horsepower, and the vehicle cannot properly react to the driver's request, and such a state may be taken into consideration during the shift determination. It should be possible to avoid meaningless upshifts if possible. Therefore, in the present control device, the appropriateness of the reaction of the vehicle is grasped from the surplus driving force and shown as a concept called control toughness, and such a concept is taken into consideration when determining the upshift. This control toughness may be any parameter as long as it can detect the amount equivalent to the surplus horsepower, but in the present invention, the running resistance, particularly the gradient resistance in it, is used for understanding. That is, as shown in FIG. 12, the slope of the traveling road is usually represented by tan θ (= height h / length 1). Based on this gradient tan θ, the membership function μ was defined for each speed stage as shown in FIGS. 13 (a) to 13 (d). In each figure, the domain (horizontal axis) has a slope (+ side indicates ascending, − side indicates downward), and the vertical axis indicates the membership value. Generally speaking, when the gradient is zero (flat), the marginal driving force decreases little, so the membership value is set to 1.0. In the 1st speed, the driving force is relatively large, so the membership value is 1.0. Approximately trapezoidal shape with a large range and a small decrease in membership value despite an increase in slope. Further, the membership function was narrowed into a triangular shape as the speed stage increased, and the triangle was sharpened especially at the highest stage. Such a control toughness fuzzy label is stored in the ROM 70d for each speed stage, and is calculated for the speed stage of the shift destination in the actual calculation. For example, in the case of rule 10, since it is planned to increase from the first speed to the second speed, the calculation will be performed for the second speed.

Rule 11 Fuzzy label Conclusion Current shift position S _{FT0 Not} shifted Acceleration change amount Δα Acceleration α Throttle change amount ΔθTH Implication of rule “Do not apply rule 10 when acceleration is not considered” It is provided as a supplement, and it is provided to prohibit the adoption of rule 10 when it is not considered that the vehicle is accelerating. That is, in fuzzy inference, the production rule with the largest evaluation value is automatically selected, so even if the evaluation value of rule 10 is not so high, the evaluation value of other rules is not so high. If is lower, rule 10 will be adopted, and as a result an inappropriate control value will be determined.
To avoid this, in this rule, one of the fuzzy labels in Rule 10, specifically, the membership function of acceleration α is a complementary set, which is a fuzzy set. That is, in the rule 10, the acceleration α is set so that the grade becomes higher when it is in the increasing direction (meaning that acceleration is in progress), whereas in this rule, when the grade is decreasing, the grade becomes higher. It is decided to grow. As a result, when the driving state is similar to the rule 10, but there is a big difference in that the vehicle is not accelerating, the grade of the acceleration α in the rule 11 becomes high, so that the adoption of the rule 10 can be prevented. Since no shift is planned in this rule, the post-shift rotation speed and control toughness will not be referred to.

Rule 12 Fuzzy label Conclusion Current shift position S _{FT0 No} gear change Acceleration change amount Δα Acceleration α Throttle change amount ΔθTH Implication of rule “If the acceleration is not saturated, no gear change” This rule is also a pair with rule 10. Then, if it is determined from the acceleration change amount Δα that the acceleration α is not saturated, the acceleration can still be performed, and it is determined that the gear shift is not performed. Therefore, also in this rule, the amount of change in acceleration Δα is complementary to that of rule 10, which prevents adoption of rule 10 in an operating state that meets this condition.

Rule 13 Fuzzy label Conclusion Current shift position S _{FT0 Does} not shift _Vehicle speed V Implications of rule "Since it is not necessary to shift up to the 2nd speed at extremely low speed, shift is not performed" It is a rule similar to rules 4-8 .

Rule 14 Fuzzy label Conclusion Current shift position S _FT0 1st speed up Throttle change amount ΔθTH Vehicle speed V acceleration α Control toughness (2nd speed) Implications of rule “When acceleration force is weak, shift up by vehicle speed” This is during slow acceleration The rule is that there is no throttle operation, the vehicle speed is stable, and the driver does not have much intention to accelerate, that is, the speed is increased by one gear to improve fuel efficiency in a gradually accelerated driving condition. Is.
As mentioned above, the acceleration is saturated during sudden acceleration, so it is possible to decide to shift up using that as an index.However, during slow acceleration, such a phenomenon is not seen, so the vehicle speed should be used as an index for shifting up. decide. As we plan to increase to second gear, we also evaluate the control toughness of second gear.

Rule 15 Fuzzy Label Conclusion Current shift position S _FT0 1st speed up Acceleration change amount Δα (from 2nd speed to 3rd speed) Control toughness CT (3rd speed) Speed after shifting Ne-SFT (3rd speed) Acceleration α Throttle change amount Δθ TH “Upshifting from 2nd speed to 3rd speed during acceleration is performed if acceleration is saturated and the control toughness of the vehicle after the shift is good.” This is the same as in Rule 10 above. However, the only difference is that the rule 10 specifies the increase from the first speed to the second speed, whereas the rule 10 specifies the increase from the second speed to the third speed. Therefore, detailed description is omitted. still,
Rules 16 and 17 are also similar to rules 11 and 12, and are complementary to rule 15.

Rule 18 Fuzzy label Conclusion Current shift position S _{FT0 No} gear change Engine speed Ne Throttle change ΔθTH Implication of rule `` If the engine speed is high enough even if the throttle operation intended to accelerate is performed, Do not shift ”When the driver pushes the accelerator pedal to intend acceleration, the control device recognizes the intention. Then, in that case, it is determined from the rotational speed whether to downshift and increase the driving force to accelerate or to accelerate without downshifting. That is, since the driving force is large when the rotation speed is high,
This is because it is possible to accelerate without changing gears.

Rule 19 Fuzzy label Conclusion Current shift position S _FT0 1st speed up Throttle change amount ΔθTH Vehicle speed V Acceleration α Control toughness CT (3rd speed) Implication of the rule Similar to rule 14 "If the acceleration force is weak, shift up depending on the vehicle speed" The difference is that the current shift position is the second speed, and as a result, the vehicle speed V is 30 km.
The point is that the membership value has risen to 1.0 at / h, and the rest is the same.

Rule 20 Fuzzy label Conclusion Current shift position S _FT0 Do not shift Throttle change amount ΔθTH Vehicle speed V Acceleration α Rule implication “If the vehicle speed is not satisfied even in the same case as above, do not upshift.” The reason why the rule 19 is not provided but is provided in the rule 19 is that the rule 14 is planned to be in the first speed, while the rule 19 is planned to be in the second speed. This is because the higher speed requires the vehicle speed to be satisfied because the driving force decreases.
Therefore, the vehicle speed V of rule 20 is set to be a complement of that of rule 19 (only on the low speed side).

Rules 21 to 26 This is similar to rules 15 to 20 and the only difference is that it is intended to shift up from the third speed to the fourth speed.

Rule 27 Fuzzy label Conclusion Current shift position S _{FT0 No} gear change Engine speed Ne Throttle change ΔθTH Implications of rule “Even if the throttle operation intended to accelerate is performed, it will accelerate when the speed is sufficiently high, so do not down.” This is similar to Rule 24, and to the same effect.

Rule 28 Fuzzy label Conclusion Throttle opening θTH 3rd speed up Throttle change amount ΔθTH Acceleration α Control toughness CT (4th speed) Rotational speed after shifting Ne-SFT (4th speed) Implications of rule “Intent to shift from acceleration to cruise” If it is clear and the controllability after shifting up is good (3rd speed), it indicates that acceleration is in progress from the acceleration α, the throttle opening θTH has been returned to the cruise opening, and If the control toughness at the 4th speed is good in an operating state in which it is expected that the rotational speed Ne-SFT after shifting at the 4th speed will also be close to 2000 rpm, the speed will be increased to the 3rd speed and shifted to the highest stage. Therefore, this rule plans when the current shift position is the first speed.

Rule 29 Fuzzy label Conclusion Throttle opening θTH No gear shift Throttle change amount ΔθTH Acceleration α Rule implications “Does not increase if the precondition is not acceleration” This is a supplement to rule 28 and is similar to rule 28 However, even if the acceleration state is not detected from the vehicle speed V, the increase is avoided. Therefore, the acceleration α of rule 29 is the complement of rule 28. With such a configuration, it is possible to avoid the inconvenience as seen in the prior art.

Rules 30 to 33 are the same as the above-mentioned rules 28 and 29, and are different only in that the planned current shift position is different.

Rule 34 Fuzzy label Conclusion Current shift position S _FT0 Third speed down Throttle opening θTH Speed after shifting Ne-SFT (1st speed) _Vehicle speed V Implication of rule “(If the throttle is fully open, the speed after shifting is extremely high. Unless otherwise) Kickdown ”This rule to the last rule 48 is a kickdown rule. In this rule, the current shift position S _FT0 is the 4th speed, it is _expected that the throttle will be depressed greatly, the post-shift rotation speed Ne-SFT at the 1st speed will not be extremely high, and the vehicle speed V When is low, the vehicle is down to the third speed and kicked down to the bottom. Rule 35
Similarly, ~ 37 is the kickdown in the case of the fourth speed, and the rule is divided according to the state of each vehicle speed, and the degree of the down is set to the second speed or the first speed depending on the difference in the throttle opening etc. To stop at.

In addition, rules 38 to 42 define kickdown when the current shift position is in the third speed, and rules 43 to 44 plan kickdown in the second speed, This is the same kind of rule as rules 35-37.

Based on the above, when the description of the fuzzy production rule search subroutine flow chart of FIG. 9 is started, the total number N of rules is read in S300. In the case of this embodiment, the total number of rules is 44.

Then, in S302, the counter that counts the rule N ₀ is initialized and the initial value is set to “n = 1” (meaning rule 1).
Set to. Subsequently, in S304, a calculation field (not shown) for the membership value μ of the RAM 70e of the microcomputer is initialized to set the membership value μ0 = 1.0.

Then, in S306, the conclusion part (gear stage number) of the n-th rule is read. Therefore, when the program is started for the first time, the conclusion "0" (without shifting) of rule 1 is read.

Subsequently, it is determined whether or not the number of gear stages read in S308 exceeds the maximum number of gear stages that can be down-shifted calculated in the flow chart of FIG. 7, and if the number of gear stages planned by the rule exceeds the maximum number of gear stages. Since there is no meaning even if that rule is selected, the process proceeds to S310 and the membership value μ0 of the rule is set to 0.0. Also, S30
If it is determined in step 8 that the maximum number of gears that can be downshifted is exceeded, then in S312, it is determined whether the conclusion of the rule is within the maximum number of gears that can be upshifted. Transition.

If the result in S312 is also affirmative, the total number L of fuzzy labels is read in S316, and the counter value 1 for counting the number of fuzzy labels is set to an initial value "1" in S316.
Set to. The number of fuzzy labels is read as 4 in Rule 1, for example.

Then, in S318, the membership value μn1 of the first label of the nth rule is read. As described above, this means the operation of applying the value actually detected or calculated from it to the domain, and reading the membership value at the position where the vertical line and the waveform raised from there are crossed, Δθ of rule 1 because it is the first time the program is started
TH = 0 (degree / 0.1s), α = 0 (km / h / 0.1s), Ne = 3000
It means to read μ11 = 1.0 for the first value in (rpm), ΔθTH. Then, in S320, the μ0 initially set in S304 is compared with the μn1 read in all steps. If μn1 is smaller than μ0, the value of μ0 is rewritten to μn1 in S322, and the label N ₀ is incremented in the next S324. Returning to S318, the membership values of all fuzzy labels are sequentially read in the same rule, and S3 is read.
After it is confirmed that the reading is completed in 26, the minimum membership value μ0 is set to the membership value μ0n of the rule n in S328. If μ0 is determined to be smaller than μn1 in S320, the process immediately jumps to S324, and if negative in S308 and 312, the membership value 0.0 is set as the minimum value of the rule. (S310). In steps S300 to 328 above, the minimum membership value of one rule is selected.

Then, in S330, the number of rules n is incremented,
Repeat until it is determined that the total number of rules is reached in the next S332, and after that, proceed to S334 to select the rule with the maximum membership value and adopt that rule (S33
6).

After that, returning to FIG. 4 again, output processing is performed according to the conclusion of the rule selected in S18, and gear shifting is performed. That is, the solenoids 36, 38 are energized / de-energized via the output circuits 74, 76 to engage or hold a desired gear.

In the present embodiment, as described above, not only the measured values of the throttle opening or the vehicle speed, but also the reaction suitability of the vehicle side to the driver's expectation is quantitatively measured from the gradient resistance and used as parameters. Based on the parameters of the expert driver, the judgment and operation of the manual transmission vehicle of the expert driver are analyzed, and a plurality of inductive control rules are set, and the optimal control value is selected by evaluating the control rules through fuzzy inference. With such a configuration, the operating states of the vehicle, including the conditions of the four circles, are captured in a multivariable manner and processed instantly, thus enabling automatic shift control similar to the judgment / operation of a trained driver in a manual transmission vehicle. Is. That is, the control using the fuzzy method enables more appropriate control similar to the manual shift operation of a human being and is restricted by the setting data as seen in the above-mentioned prior art, or the time point of shifting from the throttle opening and the vehicle speed. Is not mechanically determined. Further, by further increasing the disclosed rules, it is possible to realize the shift control corresponding to the emission countermeasure, and it is possible to more flexibly meet the shift control characteristics required by the user. In this sense, the object, structure and effect are all different from the prior art.

14 to 15 show the second embodiment of the present invention, and the difference from the first embodiment is that the number of rules is first reduced. In the first embodiment, the rule is independently defined for each current shift position, but the number is reduced by commonly defining the rule regardless of the shift position. The second point is that in the first embodiment, a complementary set is provided to deal with a rule that is not desired to be selected under a predetermined condition, whereas the membership value of the selected rule is appropriately set. It is configured to compare with a set reference value, for example, “0.5”, and if less than that, stop adopting the rule and use the previously adopted rule as it is.

These points will be briefly described below. FIG. 14 is an explanatory diagram showing the rules of the second embodiment, but the number of rules is reduced to 18 as shown. Since the individual rules are similar to those of the first embodiment, only the corresponding first embodiment rules are shown. The parameters and units used here are the same as in the first embodiment.

Rules 13-18 Rules 34-44 Next, the prohibition of rule selection will be described. FIG. 15 is a flowchart showing a rule search subroutine in the second embodiment.
9 is a chart in which the flow chart of FIG. 9 of the first embodiment is partially changed (thus, only the changed portion is shown). Explaining below, in S332, it is determined that all the rules have been searched, and in S334, the rule having the maximum membership value is selected. Then, in S334a, the membership value μ0 of the selected rule is set to a predetermined reference value μref, for example. Compared with the above-mentioned 0.5, if it exceeds the rule, the rule is adopted in S336, and if it does not exceed the rule, the process proceeds to S334b to hold the current shift.

In the case of the second embodiment, as described above, in addition to the advantages of the first embodiment, the memory capacity of the microcomputer can be reduced due to the reduction of the control rules, and the calculation time can be shortened. It has the advantage that The rule selection prohibition itself is the same as that of the first embodiment in that the rule selection can be prohibited even though the evaluation value is absolutely low, but the rule selection is relatively high. Is also the same.

(Advantages of the Invention) A control device for an automatic transmission according to the present invention is a driving state detecting means for detecting a driving state of an engine or a vehicle including an engine speed, a throttle opening, a throttle opening speed, a vehicle speed and a vehicle acceleration, For all gears that can be shifted from the current gear,
Operating state prediction that predicts changes in operating conditions that would occur if the gear position is changed, including changes in marginal driving force and engine speed that are obtained through running resistance calculated based on the road surface inclination angle Means, membership function setting means for setting a membership function in a fuzzy set in advance for the driving state to be detected and the driving state change to be predicted, and a plurality of the detected value and the predicted value and the set membership function A fuzzy production rule is selected to determine a speed change gear speed, and a drive means for driving the speed change mechanism according to the output of the speed change gear speed determining means; The speed change gear determining means is configured to prohibit selection of a specific fuzzy production rule in a predetermined operating state. Therefore, by grasping the driving state of the vehicle including the conditions of four circles with multiple variables and processing it instantaneously through fuzzy reasoning, the judgment / action similar to the gear change judgment / operation performed by the expert driver in the manual transmission vehicle can be performed. This can be realized during the control, and the shift can be determined more accurately as compared with the conventional device. Also,
By prohibiting the selection of a specific fuzzy production rule in a given operating condition, an inappropriate control value will not be determined. Therefore, as it is not seen in the prior art, it is not necessary to mechanically determine the shift time point from the throttle opening and the vehicle speed based on the preset shift diagram, so that the shift control that responds to the ever-changing driving state is realized. You can do it. Further, by appropriately determining the shift rule, it is possible to realize shift control corresponding to the emission countermeasure or shift control capable of individually responding to the shift characteristics required by each user.

[Brief description of drawings]

FIG. 1 is an explanatory view showing an embodiment of the present invention, FIG. 2 is a schematic view showing an overall configuration of a control device for an automatic transmission according to the present invention, and FIG. 3 is a block showing details of the control unit. 4 and 5 are flow charts showing a shift main routine which is an operation of the unit, FIG. 5 is a flow chart showing a shift command value determining subroutine therein, and FIG. 6 is explaining parameter calculation therein. Explanatory drawing, FIG. 7 is the fifth
FIG. 8 is a flow chart showing a subroutine for calculating the number of revolutions after shifting, FIG. 8 is an explanatory view showing an example of the calculation, and FIG. 9 is a flow showing a fuzzy production rule search subroutine in the flow chart of FIG.・ Chart, FIG. 10 is an explanatory diagram showing the rule, FIG. 11 is a driving force diagram showing the premise of control toughness used in the rule, and FIG. 12 is an explanatory diagram of gradient resistance represented by control toughness. 13 (a) to 13 (d) are explanatory diagrams showing the control toughness in which a membership function is given for each shift position, and FIG. 14 is an explanation showing the fuzzy production rule used in the second embodiment of the present invention. Figure and Fifteenth
The figure is a flow chart of the main part of the rule search subroutine in the second embodiment. 10 ...... Internal combustion engine body, 16 ...... Throttle valve, 18 ...... Engine output shaft, 20 ...... Transmission, 22 ...... Torque converter, 24 ...... Main shaft, 26 ...... Counter shaft, 30 ...... Oil passage, 32 , 34 …… Shift valve, 36,38 …… Electromagnetic solenoid, 42 …… Differential device, 46 ……
Rear wheel, 50 ... Throttle sensor, 52 ... Crank angle sensor, 54 ... Inclination angle sensor, 56 ... Vehicle speed sensor, 60 ... Gear change control unit, 62 ... Range selector switch, 64 ...
… Shift position switch, 70… Micro computer

Claims (4)

[Claims] Claims: 1. a. Operating state detecting means for detecting the operating state of an engine or vehicle including engine speed, throttle opening, throttle opening speed, vehicle speed and vehicle acceleration; b. For gears, predict changes in operating conditions that would occur if a shift is made to the gears, including changes in surplus driving force and changes in engine speed that are obtained from running resistance calculated based on the road surface inclination angle. Driving state predicting means, c. Membership function setting means for setting a membership function in a fuzzy set in advance for the driving state to be detected and the driving state change to be predicted, d. The detected value and the predicted value and setting A shift gear determining means for determining a shift gear by selecting one of a plurality of fuzzy production rules from the determined membership function, and e. Drive means for driving the speed change mechanism in accordance with the output of the speed change gear stage determining means, and the speed change gear stage determining means prohibits selection of a specific fuzzy production rule in a predetermined operating state. Control device for automatic transmission. 2. The control device for an automatic transmission according to claim 1, wherein the prohibition is performed by providing a complement to the membership functions of the plurality of fuzzy production rules. 3. The prohibition compares the membership value of the selected fuzzy production rule with a reference value,
3. The automatic transmission control device according to claim 2, wherein when the number of rules is not exceeded, the selected rule is discarded. 4. The automatic system according to claim 1, wherein the fuzzy production rule is created based on a driver's operation / judgment made in a vehicle with a manual transmission. Transmission control device.