JP7795985B2 - Design support device and design support method - Google Patents

Description

The present invention relates to a design support device and design support method that support the design of industrial machinery with multiple design variables.

When designing industrial machinery, multiple design variables must be set appropriately to ensure that the required performance is met. In such cases, changing any one design variable may result in the required performance not being met, making it necessary to change design variables other than the one you want to change. The more design variables there are, the more difficult it becomes to determine which other design variables need to be changed to meet the required performance, resulting in numerous trial and error processes.

This has led to a demand for design support devices that can assist in setting design variables. As part of this effort, for example, Patent Document 1 proposes a search device that efficiently searches for valid ranges that represent the range of design variables that satisfy various performance conditions and constraints.

JP 2020-64479 A

However, with the device described in Patent Document 1, even if the design variables to be changed and their corresponding change values are given, it is not possible to suggest other design variables that should be changed to meet the required performance.

The greater the number of design variables set for industrial machinery, the greater the number of combinations of other design variables that must be changed to meet required performance, and the greater the time and cost required for trial and error. Therefore, there is a demand for design support devices and methods that reduce trial and error as much as possible and improve design efficiency by enabling rapid design changes.

In view of the above, the present invention provides a system including: a first design variable calculation unit that calculates a value of each design variable based on a relationship between a required performance of an industrial machine and a plurality of design variables in the required performance; an output unit that outputs the value of each design variable calculated by the first design variable calculation unit; and an identification unit that identifies an arbitrary number of design variables from among the plurality of design variables, wherein the first design variable calculation unit calculates the value of the design variable identified by the identification unit based on a range of the required performance of the industrial machine, and further calculates values of design variables other than the design variables identified by the identification unit based on the value of the design variable calculated by the first design variable calculation unit and the range of the required performance of the industrial machine. and a processing unit that processes the design variables, wherein the processing unit, when allowed to be within a region of a predetermined width for a first design point determined by the first design variable identified by the identifying unit and the second design variable calculated by the second design variable calculation unit, presents the first design point, along with a region of a predetermined width, on a plane that displays a required performance range, with design variables on the vertical and horizontal axes on the output unit, and wherein, when a plurality of values of the second design variable that bring the required performance of the industrial machine into the required performance range are detected as a result of the search, the processing unit selects the side that can widen the region of a predetermined width.

Further, in the present invention, there is provided a design support method for a design support device including a first design variable calculation unit, an output unit, an identification unit, a second design variable calculation unit, and a processing unit, wherein the first design variable calculation unit calculates a value of each design variable based on a relationship between a required performance of an industrial machine and a plurality of design variables in the required performance, the output unit outputs the value of each design variable calculated by the first design variable calculation unit, the identification unit identifies an arbitrary number of design variables from among the plurality of design variables, the first design variable calculation unit calculates the values of the design variables identified by the identification unit based on a range of the required performance of the industrial machine, and the second design variable calculation unit outputs the values of the design variables calculated by the first design variable calculation unit and a design support method for a design support device, characterized in that: based on the range of the required performance of the industrial machinery, the values of design variables other than the design variable specified by the specifying unit are calculated; the processing unit, when allowed to be within a region of a predetermined width for a first design point determined by the first design variable specified by the specifying unit and the second design variable calculated by the second design variable calculating unit, displays the first design point together with a region of a predetermined width on a plane displaying the required performance range, with design variables on the vertical and horizontal axes on the output unit; and when, as a result of the search, the processing unit detects multiple values of the second design variable that bring the required performance of the industrial machinery into the required performance range, the processing unit selects the side that can widen the region of the predetermined width.

This invention improves design efficiency by comparing the design variables to be changed and their values with a pre-defined required performance range, and then proposing other design variables that should be changed to meet the required performance.

1 is a diagram showing an example of the configuration of a design support apparatus according to an embodiment of the present invention; FIG. 1 is a diagram showing the relationship between design variables of an industrial machine and a required performance range. FIG. 10 is a diagram for explaining a proposal of a design variable B as another design variable to be changed in response to a change in a design variable A in order to satisfy a required performance range. FIG. 10 is a graph showing a design variable B and a performance value at a change point. FIG. 10 is a graph showing a design variable B and a performance value at a change point. FIG. 1 is a diagram illustrating the concept of design constraint range. FIG. 10 is a diagram showing the concept of proposing multiple design constraint ranges. A diagram showing the idea of proposing other candidates. FIG. 10 is a diagram showing the concept of proposing another candidate that is closest to the changed value of design variable A. FIG. 10 is a diagram showing the concept of proposing another candidate that is closest to the changed value of design variable A. 1 is a diagram showing the relationship between design variables B and C in an industrial machine having design variables A, B, and C. 1 is a diagram showing the relationship between design variables B and C in an industrial machine having design variables A, B, and C.

An embodiment of the present invention will now be described with reference to the drawings.

First, an example of the basic configuration of a design support device according to an embodiment of the present invention will be described with reference to Figure 1. The design support device 100 shown in Figure 1 is composed of a computer device, and generally includes external input devices such as a keyboard and output devices such as a monitor, and is connected via a bus to ROM, RAM, a processing unit (CPU) that performs calculations, a database, etc. Note that Figure 1 shows processing functions executed by software processing in the processing unit (CPU), and does not depict the hardware elements mentioned above.

When implementing the design support device according to the present invention, it is preferable to use a personal computer that includes an input unit for inputting at least the design variables that determine the required performance of the industrial machinery, and a display unit for outputting the calculation results.

Specific examples of use of a design support device according to an embodiment of the present invention will be described from Example 2 onwards. Example 1 will explain the required performance range calculation unit 101, a basic element that a design support device should have. The required performance range calculation unit 101 can be prepared in advance of actual use and provided as, for example, a required performance range database, or it can be calculated each time the device is used. Here, the explanation will be given with the assumption that it will be provided as a required performance range database.

The required performance range database stores the relationships shown in Figure 2, and provides required performance range 1 for each required performance against design variables A and B on the vertical and horizontal axes. Specifically, to design and manufacture industrial machinery, it is necessary to satisfy multiple required performance (constraints), such as "The metal temperature at part XX must be below T1 degrees," "The pressure at part XX must be above P1," and "The vibration at part △△ must be below G1."

In this invention, the influencing factors (influencing factors) of these required performance (constraints) are already known. For example, it is known that "the metal temperature of part XX is determined by design variables A and B," "the pressure of part XX is determined by design variables A and C," and "the vibration of part △△ is determined by design variables D, E, and F."

Furthermore, the relationship between design variables and required performance is also known from prior analysis. An example of this relationship is required performance range 1 in Figure 2, where, for example, within this range, "the metal temperature at part XX can be kept below T1 degrees," "the pressure at part XX can be kept above P1," and "the vibration at part △△ can be kept below G1."

In the present invention, all of this is achieved through prior analysis and configured as a required performance range database, or the required performance range calculation unit 101 may receive required performance and design variables as input each time the system is put into actual use, and calculate the corresponding design variables and required performance to provide as output.

Required performance range 1 may be defined as the point cloud information in Figure 2, or may be expressed as a function, and any method of realization may be used. The number of influencing factors of the required performance (number of design variables) may be three or more, but for the sake of explanation, we will use two, and explain on the two-dimensional plane of Figure 2. However, a certain design variable may be an influencing factor for multiple required performances, and changing it may not necessarily result in an improvement in both required performances.

In the first embodiment, it was explained that the required performance range calculation unit 101 should be provided with, for example, a required performance range database. In the second embodiment of the present invention, when configuring a design support device using a computer device, it is proposed that the following calculation processing units should be provided as basic functions that should be provided in the calculation unit CPU of the computer device.

The first of these is a function to access the required performance range database 101, and includes a design variable calculation unit that calculates the value of each design variable based on the relationship between the required performance of the industrial machinery and multiple design variables in that required performance. Note that the design variable calculation unit may be provided with multiple functions, one for each of multiple design variables.

The second feature is that when accessing the required performance range database 101, it is equipped with an identification unit that identifies any number of design variables from among multiple design variables, that is, that sets what to access.

The third feature is that it is equipped with a processing unit that processes the values of design variables obtained by accessing the required performance range database 101 and processes them into meaningful information.

In the third and subsequent examples, we will explain the application and practical use of a design support system realized by combining the basic functions described above. Figure 1 shows an example of the configuration of a design support system according to an embodiment of the present invention, and shows the processing functions executed by software processing in the processing unit (CPU) of the computer system.

The design support device 100 is composed of a required performance range calculation unit 101 that calculates the required performance range of an industrial machine having multiple design variables; a setting unit 102 that sets any number of design variables from the multiple design variables of the industrial machine; a first change unit 103 that changes the values of the design variables set by the setting unit 102 based on the required performance range of the industrial machine calculated by the required performance range calculation unit 101; a second change unit 104 that changes the values of design variables other than the design variable set by the setting unit 102 based on the range of values of the design variables changed by the first change unit 103 and the required performance range; a calculation unit 105 that calculates the range of values of the design variables changed by the first change unit 103 and the second change unit 104; and an output unit 106 that outputs the range of values of each design variable of the industrial machine.

In comparison with the basic functions described in Example 2, the design variable calculation unit is the first change unit 103 and the second change unit 104, the identification unit is the setting unit 102, and the processing unit is the calculation unit 105.

Note that here, it is assumed that the required performance range calculation unit 101 is provided as a required performance range database that has been prepared and configured in advance, and that "metal temperature of part XX" is specified in advance as the required performance, and that Figure 2 is displayed on an output device such as a monitor (not shown) via the output unit 106 as required performance range 1, where "metal temperature of part XX is T1 degrees or less."

Figure 2 shows a plot of combinations of design variables A and B. Here, design variables A and B are plotted on the vertical and horizontal axes, and the range of combinations of design variables that satisfy the performance required by the required performance range calculation unit 101 is shown as required performance range 1. In other words, it shows that any combination of design variables A and B that exists within this required performance range 1 can satisfy the requirements.

In this example, we assume that there is a positive correlation between the design variable combination position (design point 20) so that it falls within required performance range 1, such that increasing the value of design variable A requires increasing design variable B as well. Hereafter, for ease of explanation, we will use only two design variables, A and B, but there can be three or more explanatory variables, and there is no upper limit to the number.

When the required performance range calculation unit 101 described above is implemented online, a simulation or experiment is performed for each plot point to calculate the performance for each combination of design variables. Here, a three-dimensional design space is created based on design variable A, design variable B, and the calculated performance, and required performance range 1 is calculated. The required performance range calculation unit 101 may be created automatically, with the range stored in a database, and the database may be referenced during operation, or the required performance range may be calculated each time.

For example, one method is to create contour lines of performance values for the design space and use those contour lines as thresholds for the required performance to determine the required performance range 1. At this time, with the aim of shortening the time required for multiple simulations or experiments, another method is to use a combination of multiple design variables as explanatory variables, and the performance obtained based on those explanatory variables as the target variable, and use a regression model obtained by machine learning techniques or a response surface obtained by response surface methodology as the design space 2.

In this invention, it is assumed that a required performance range 1, determined by a combination of design variables that can satisfy the required specifications, is determined in advance.

Figure 3 is a diagram to explain how design variable B is proposed as another design variable to be changed in order to satisfy required performance range 1 in response to a change in design variable A. In Figure 3, the current design point 20, where design variables A and B are both 0, is used as the reference, and design point 20 is assumed to be within required performance range 1.

According to this diagram, the initial design point 20 is within the required performance range 1 and fully meets the required specifications, but there may be circumstances in which you wish to change this setting variable A. This may be the case when you want to find a more suitable point, or when you want to apply design variables defined for another industrial machine to a similar industrial machine to check its performance.

When a designer attempts to make a design change to reduce design variable A for such design point 20, the designer using the design support system sets design variable A as the design variable to be changed using the setting unit 102 in Figure 1. The designer also specifies the change difference value for design variable A themselves, or the change difference value is automatically set in the first change unit 103. In this case, the change difference value for design variable A is minus 2 in the example in Figure 3, which is change point 21 in Figure 3, and this state is displayed on the monitor as it is changed.

In this case, change point 21 falls outside required performance range 1, and the design is invalid. In this regard, as was the case with conventional technology, if required performance range 1 is unknown, trial and error is required, such as changing the setting value for design variable A and changing the setting value for other design variable B. In this case, the more design variables there are, the more trial and error there is, and as a result, it takes a huge amount of time and cost to determine the appropriate change point 21.

On the other hand, if required performance range 1 is known by required performance range calculation unit 101 and is displayed on a monitor or the like as shown in Figure 3, the design will be valid even if the value of design variable A set by the designer remains unchanged. In other words, by proposing a change to another design variable that needs to be changed to bring change point 21 within required performance range 1, in this case design variable B, the design will fall within required performance range 1 and the design will be valid.

For example, in the display example in Figure 3, the factors influencing the required performance "metal temperature of part XX" are design variables A and B, and since design variable B, other than the initially changed design variable A, is displayed on the vertical axis, it is suggested that design variable B be changed, and that by changing the amount by -2, it will fall within the required performance range of 1. This is change 22 in Figure 3, and the suggestion is made by changing and displaying this state on the monitor.

Figure 4 shows a graph of design variable B and performance values at change points. The horizontal axis represents the magnitude of design variable B, and the vertical axis represents the performance at this magnitude, showing the range of design variables where the performance characteristics (design space 2) are equal to or greater than the threshold value of the required performance. As shown in Figure 4, by comparing change point 21 with design space 2, the performance value at change point 21 and the gradient 40 in design space 2 are calculated. Based on the difference between the performance value at change point 21 and the threshold value of the required performance, it is possible to propose the direction of change to design variable B according to gradient 40.

Figure 5 also shows a graph of design variable B and performance values at change points. As shown in Figure 5, by keeping the value of design variable A fixed and setting multiple design variables B, it is possible to propose the direction of change for design variable B based on the performance values of multiple change points 210.

In this way, the output unit 105 proposes values for design change B so that the design falls within required performance range 1, and based on the proposal results, the designer combines the values of design variable A and design variable B to determine the post-change design point 22.

The above processing in Example 3 is realized by the functions of the setting unit 102, first change unit 103, and required performance range calculation unit 101 in Figure 1 working together to display externally on a monitor.

In Example 3, the design variables were specific numerical values without a range, and only changes to the design points were discussed. However, in Example 4, we will consider a case where the design variables have a range and are represented as a design domain. As a specific example of a design domain, in actual industrial machinery, some constraints may be imposed on the design points. For example, if the design variable is a dimension, the dimensional tolerance would be that constraint.

In Example 4, in addition to the functions of the setting unit 102, first change unit 103, and required performance range calculation unit 101 shown in Figure 1, the functions of the second change unit 104 and calculation unit 105 are also used.

Figure 6 is a diagram illustrating the concept of the design constraint range. In Example 3, the design variables were explained as being constant values (set points), but when dimensional tolerances and the like are taken into consideration, there is a range (design constraint range) for the combination of design variables, and they should be considered in a region centered on the set point. In Example 2, the combination of design variable A and the design variable was a constant point indicated by 20, but in Example 4, it is considered as a range or region 201 with a range of dimensional tolerances and the like centered on 20.

In Figure 6, it is assumed that design constraints (design constraint range) 201 are imposed on design variables A and B for design point 20, which is determined by the current combination of design variables A and B. In other words, when the setting unit 102, first change unit 103, and required performance range calculation unit 101 are displaying the point in Figure 3, if it is confirmed that design constraints exist (it is assumed that the design constraints are known and stored in advance), the second change unit 104 changes design point 20 to area 201 that includes the design constraints and displays it.

The second change unit 104 also applies the second change, which results in a similar area display, to the modified design point 22. In this case, the process up to determining the modified design point 22 is the same as in Example 2, so a description thereof will be omitted. Once the modified design point 22 has been determined, the second change unit 104 first sets the design constraint 220, which has the same range as the design constraint 201, to the modified design point 22.

Furthermore, the calculation unit 105 then functions, comparing the required performance range 1 with the design constraints 220, and proposing new design constraints 221 if the design constraints 220 should be expanded or reduced. Based on the results of this proposal, the designer can determine the new design constraints 221.

According to Example 4, a new design constraint range can be set for the changed design point 22. This means, for example, that the dimensional tolerance of the changed design point 22 can be set more leniently than the conventional dimensional tolerance, thereby achieving the effect of producing an inexpensive product.

Figure 7 shows the concept of proposing multiple design constraint ranges. In Figure 7, we assume required performance range 2, in which the required design range has a complex shape. The present invention can be implemented regardless of the shape of the design range.

In Figure 7, if a design change is made to reduce design variable A for the current design point 20, change point 21 will fall outside required performance range 2, and the design will not be viable. In this case, a change to design variable B will be proposed using the methods described in Examples 2 and 3, but as shown in Figure 7, multiple design points may be proposed.

In this case, the post-change design point 23 and the post-change design point 24 are simultaneously output by the output unit 106, and the design constraints 230 and 240 that can be set at each design point are also output.

Based on the multiple proposals displayed on the monitor, the designer can select post-change design point 23 and design constraint 230 if they want to reduce the change value of design variable B even if it reduces the range of the design constraint; or select post-change design point 24 and design constraint 240 if they want to increase the range of the design constraint even if it increases the change value of design variable B.

It is also possible to select the post-change design point 24 and set the design constraints within the same range as the current design constraints 201. This provides guidelines for selecting the post-change design point and allows the designer to explain the reasons for their selection.

Figure 8 shows the concept of proposing other candidates. Using Figure 8, we will explain a case where, in the required performance range 2 used in Example 4, for the initial point 20 with a region of 200, the change point 21 does not fall within the required performance range 2 even when the other design variable B is changed (moved in the vertical direction).

At this time, at change point 21, the change value for design variable A is about -3.5, which is too large. Even if design variable B is changed to search for a region that matches required performance range 2 along the vertical axis, it does not fall within required performance range 2, and the design is invalid. In this case, with the intention of reducing design variable A, a value that falls within required performance range 2 is searched for by reducing design variable A (for example, by -2.5). As a result of the search, changed design point 25 is proposed.

When the required performance range 2 described in Example 6 is not met, a method for setting the design variable A so that it is as close as possible to the changed value will be explained using Figure 9.

In Figure 9, first, only the value of design variable A is changed (moved along the horizontal axis) to search for a search point 211 that overlaps with required performance range 2. Then, a gradient 41 is calculated at search point 211 along the frame of required performance range 2. According to gradient 41, search point 211 is moved in a direction approaching change point 21, and the nearest point 212 that is closest to change point 21 is determined. Finally, a changed design point 26 and design constraint 260 are proposed that fall within required performance range 2 even when the same range as design constraint 200 imposed on the current design point 20 is set. These processes are performed by the calculation unit 105.

Here, we have given an example of a search method using a gradient 41, but another method is to generate multiple change points 213 centered around change point 21, as shown in Figure 10, and then determine the nearest point 212 from among them.

Examples 1 to 7 have been explained using two design variables, but they can also be applied to three or more design variables. Here, assume an industrial machine has design variables A, B, and C.

In this case, if design variable A is reduced, it will fall outside the required specification range 1, as shown in Figure 3, and a design change to design variable B will be proposed. Here, we will assume a four-dimensional design space consisting of design variables A, B, C, and performance, and explain this using Figure 11, which is a graph showing a two-dimensional space shown only with design variables B and C.

In this case, to stay within required performance range 1, it is sufficient to change design point 27, which only changes design variable B, and there is no need to change design variable C. In this case, by excluding design variable C from the proposal, the designer can focus only on the design variables that need to be changed, further improving design efficiency.

In an industrial machine with design variables A, B, and C as described in Example 8, if a change in design variable A requires changes to both design variables B and C as shown in Figure 12, a more flexible design is possible by comparing the possible range 28B of design variable B with the possible range 28C of design variable C within required performance range 1 and proposing to prioritize the smaller design variable value. In Figure 12, because range 28B is smaller than range 28C, the proposal is to determine design variable B first, followed by design variable C.

In Figure 1, the output unit 106 provides output to, for example, a monitor, and it is recommended that the monitor screen display the diagrams shown in Figures 2 to 12. These diagrams show multiple design variables and the correlation characteristics between these design variables and the required performance as a required performance range, and show the position on the required performance range of the correlation characteristics when some of the design variables are changed. The relationship with this required performance range is also shown when other design variables are changed.

1: Required specification range 2: Design space 20: Current design point 21: Change points 22 to 28: Changed design points 100: Design support device 101: Required performance calculation unit 102: Setting unit 103: First change unit 104: Second change unit 105: Calculation unit 106: Output unit

Claims (8)

a first design variable calculation unit that calculates a value of each design variable based on a relationship between a required performance of the industrial machine and a plurality of design variables in the required performance;
an output unit that outputs the values of the design variables calculated by the first design variable calculation unit;
an identification unit that identifies an arbitrary number of design variables from among the plurality of design variables;
Equipped with
the first design variable calculation unit calculates values of the design variables identified by the identification unit based on a range of required performance of the industrial machine;
moreover,
a second design variable calculation unit that calculates values of design variables other than the design variables identified by the identification unit, based on the values of the design variables calculated by the first design variable calculation unit and the range of the required performance of the industrial machine;
a processing unit for processing the design variables;
Equipped with
the processing unit, when a first design point determined by the first design variable identified by the identifying unit and the second design variable calculated by the second design variable calculating unit is allowed to be within a region of a predetermined width, presents the first design point together with the region of a predetermined width on a plane that displays a required performance range, with design variables on the vertical and horizontal axes, on the output unit;
the processing unit, when detecting a plurality of values of the second design variable that make the required performance of the industrial machine fall within the required performance range as a result of the search, selects the side that can widen the region of the predetermined width . 2. The design support device according to claim 1 ,
the first design variable calculation unit and the second design variable calculation unit calculate the value of each design variable so that the range of the value of each design variable is maximized within a range of required performance of the industrial machine. 2. The design support device according to claim 1 ,
the processing unit, when the value of the first design variable is changed and the required performance of the industrial machine falls outside the required performance range, presents on the plane the value of a second design variable that brings the required performance of the industrial machine into the required performance range, and a second design point determined by the first design variable, together with an area of a predetermined width. 4. The design support device according to claim 3 ,
a processing unit for processing the second design point on the plane, the processing unit adjusting the area of the second design point to a size different from that of the area of the first design point; 2. The design support device according to claim 1 ,
the processing unit fixes the value of the first design variable after the change, and variably adjusts the value of the second design variable to search for a value of the second design variable that contributes to bringing the required performance of the industrial machine into the required performance range. 2. The design support device according to claim 1 ,
the processing unit, when fixing the value of the first design variable after the change and variably adjusting the value of the second design variable, is unable to find a value of the second design variable that contributes to bringing the required performance of the industrial machine into the required performance range, changes the value of the first design variable after the change and searches again. 2. The design support device according to claim 1,
A design support device comprising a display unit that displays the information output by the output unit. A design support method in a design support apparatus including a first design variable calculation unit, an output unit, an identification unit, a second design variable calculation unit, and a processing unit, comprising:
the first design variable calculation unit calculates a value of each design variable based on a relationship between a required performance of the industrial machine and a plurality of design variables in the required performance;
the output unit outputs the values of the design variables calculated by the first design variable calculation unit;
the identifying unit identifies an arbitrary number of design variables from among the plurality of design variables;
the first design variable calculation unit calculates values of the design variables identified by the identification unit based on a range of required performance of the industrial machine;
the second design variable calculation unit calculates values of design variables other than the design variables identified by the identification unit, based on the values of the design variables calculated by the first design variable calculation unit and the range of the required performance of the industrial machine;
the processing unit, when a first design point determined by the first design variable identified by the identifying unit and the second design variable calculated by the second design variable calculating unit is allowed to be within a region of a predetermined width, presents the first design point together with the region of a predetermined width on a plane that displays a required performance range, with design variables on the vertical and horizontal axes, on the output unit;
When a plurality of values of the second design variable that make the required performance of the industrial machine fall within the required performance range are detected as a result of the search, the processing unit selects a side that can widen the region of the predetermined width.
A design support method in a design support device.