JPH0640927B2 - Monitoring control device for water supply block - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention relates to monitoring and control of flock formation in a flock formation pond (mixing pond) of a water purification plant, and is suitable for quantitatively grasping the state of flocculation formation. The present invention relates to a block monitoring control device to which recognition technology is applied.

[Conventional technology]

In a water purification plant, a flocculant is injected into the raw water taken up to flocculate suspended matter to form flocs which are flocculates, and the flocs are sedimented and removed. Specifically, after injecting the coagulant in the rapid mixing pond, it is introduced into the flocks forming pond and the flocs are formed while gently stirring. Raw water flowing out of the floc formation pond is guided to the sedimentation pond, where the floc is sedimented and suspended substances are removed. The fine flocs that have not settled in the settling basin are further removed in the filtration basin.

In this way, when water treatment is performed, if the good flocs are not formed in the flocculation pond, the clogging of the filtration pond will be accelerated. Therefore, it is essential to monitor whether a good flock is formed. Conventionally, the state of flock formation is monitored by visual observation several times a day. For this reason, continuous and quantitative monitoring is not possible, and it is inevitable that the discovery of an abnormal situation in which no block is formed will be delayed or that countermeasures will be delayed. In order to solve such a problem, a method of monitoring the shape and size of the flocks by image processing as disclosed in Japanese Patent Laid-Open No. 54-143296 has been proposed. Specifically, from a block image taken by an industrial camera or the like, a portion (pixel) brighter than a predetermined brightness (threshold value) is recognized as a “1” level, and conversely a portion (pixel) darker than a predetermined value is detected. ) Is recognized as a level other than the block as a “0” level. In this way, the block image is binarized and image processing is performed to monitor the state of block formation.

[Problems to be solved by the invention]

In the above-mentioned conventional technique, the portion where the brightness of the flock image is brighter than the threshold value is regarded as the flock, and conversely, the portion where the brightness is darker than the threshold value is regarded as the background, which is binarized. It is an attempt to grasp and monitor. However, the actual visual inspection by the maintenance manager at the water purification plant is not only looking at the particle size or particle size distribution, but also how a sticky flock is formed is an important criterion as a factor. I'm running. That is, in the case of a floc that looks slippery, it has been obtained as a long-term operation know-how of the maintenance manager that the settling property in the sedimentation pond is good even if the particle size is a little small. The "slipping visible flocks" here means a state where the flocculation flocks and the surrounding water containing suspended suspended solids have a high contrast, and there is little suspended suspended solids and the water is transparent. Is shown. That is, it means that the suspended solids in the raw water form a floc with a good flocculation and sedimentability due to the coagulant, and the suspended suspended solids are small.

As described above, the monitoring of the floc formation situation in the water treatment plant involves the problem that it must be carried out from the two aspects of the coagulated floc particle size and how the floc looks sticky. However, the conventional technique using the above-described image processing has a drawback in that it cannot be said that sufficient flotation monitoring is performed because it is monitored only from the aspect of the shape and size (particle size) of the flocks.

An object of the present invention is to provide a monitoring control device for a clean water flock capable of simultaneously detecting the particle size of the floc and the stickiness of the flock.

[Means for solving problems]

MEANS TO SOLVE THE PROBLEM In order to achieve the above-mentioned object, first of all, the fact that a human can visually see through the eyes means that there is a high contrast between the flocs of flocculation and the surrounding water containing floating suspended matter, that is, the surroundings of the flocks. It is noted that the water in the field has low suspended solids, is transparent, and has low turbidity.

The above-mentioned purpose is to binarize the image of the inside of the flocks forming pond captured by an industrial television camera with a certain threshold value, recognize the flocks, and obtain the particle size and particle size distribution, and at the same time, make the turbidity of parts other than the flock This is achieved by obtaining a brightness level as a degree index and comparing this brightness level with the brightness level of the block (the contrast ratio between the block and the periphery of the block).

[Action]

The turbidity index of the background can be obtained by removing the flock part from the captured image of the flock formation pond and detecting the brightness level of the surrounding background. By comparing this brightness level and the brightness level of the flocks, it is possible to quantitatively detect the adhesiveness when viewed with the eye. It becomes possible to carry out comprehensive monitoring for the formation of flocs with excellent sedimentation in a sedimentation pond.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 1, raw water is introduced into the rapid mixing basin 100 while an inorganic coagulant such as a coagulant (polyaluminum chloride) or aluminum sulfate is supplied by a coagulant injection pump 120. Although not shown, an alkaline agent such as calcium hydroxide or sodium carbonate is injected to promote the formation of flock. Stirring blade 140 in rapid mixing pond 100
Are stirred by the stirrer 130. The water stirred with the coagulant injected is the coagulation tank (hereinafter referred to as the floc formation pond) 1
Guided to 50. Flotch formation pond 150 is three ponds 15
0A, 150B and 150C, which are partitioned by straightening walls 160A and 160B having a plurality of holes on the wall surface. Flot formation ponds 150A, 150B and 150C
Wooden stir paddles 170A, 170B and 1 each
70C is installed and slowly rotating around 1 to 10 rpm (paddle peripheral speed = 0.15 to 0.8 m / s). 170C of stirring paddles in 150C of flock formation pond
An imaging device 180 such as an underwater camera is installed at a position where it does not collide with. The water containing the flocs grown in the flocks forming pond 150 is guided to the sedimentation pond 190, where the flocs are settled and removed. The supernatant water from which the flocs have been removed flows into the filter pond 195. Sedimentation pond 190 in filtration pond 195
At this time, the fine flocs remaining without being sedimented and removed are filtered and removed. The water that has passed through the filtration basin 195 is supplied through a drainage basin and a reservoir, which are not shown.

The analog electric signal of the grayscale image of the water containing the aggregate obtained from the imaging device 180 is transmitted to the image recognition means 300. The image recognition means 300 includes an image processing start timer 350,
It is composed of a grayscale image information storage means 310 and a binarization means 320. The image processing start-up timer 350 periodically activates the image processing once every 5 minutes to 1 hour. The grayscale image information storage means 310 stores the grayscale image of the water containing the aggregate obtained by the image pickup device 180 in a grayscale image memory (not shown in FIG. 1) as a digital value. The grayscale image memory is shown in FIG. The resulting image has a length of 25
6, the horizontal is divided into 256 (the divided portion is referred to as a pixel), and the grayscale image memory corresponds to this.
The storage area of 56 × 256 has a brightness (luminance) value g (i, j): i = 1 to 2 in each storage area of the i-th row and the i-th column.
56, j = 1 to 256 are stored. Here, the horizontal array is i and the vertical array is j. The binarization means 320 and the grayscale image information storage means 310 receive the output results of the grayscale images and binarize them.

The particle size distribution calculation means 400 receives the output result of the binarization means 320, calculates the floc particle size distribution, obtains the volume assuming that the flock is a sphere, and converts it into the volume concentration distribution, and the result is the volume concentration distribution. It is stored in the memory 410. On the other hand, it receives the outputs of the image brightness distribution calculation unit 500 and the grayscale image information storage unit 310, calculates the brightness distribution (histogram) of the entire screen, and stores the calculation result in the brightness distribution memory 510. 55
Reference numeral 0 denotes a background luminance calculating means which receives the result of the luminance distribution memory 510 and calculates the luminance of the portion other than the flocks containing the suspended suspended matter, that is, the background, and the calculated result is the background luminance memory 560
To store. A block brightness calculation unit 650 receives the result of the brightness distribution memory 510, calculates the brightness of the block, and stores the calculation result in the block brightness memory 660. Background / block brightness difference (contrast) calculation means 7
00 is a background luminance memory 560 and a block luminance memory 66.
The brightness difference between the background and the block is calculated from the result of 0 and stored in the background / brightness difference memory 710. Recognition end determination means 800
Determines whether or not the number of times of recognition of the block image has ended a predetermined number of times. If not, the gray-scale image information storage means 310 at that point in time relates to the gray-scale image information storage means 310 described above for the gray-scale image captured by the imaging device 180.
The image processing operation from the particle size distribution calculating means 400 to the background / float brightness calculating means 700 is repeated. The recognition end determination means 800 determines whether or not this repeated operation has been repeated a predetermined number of times (for example, 10 times). When one image processing takes 20 seconds, it takes 200 seconds by repeating 10 times.

When it is determined that the image processing of a predetermined number of times is completed, the flocc log average particle diameter is calculated from the volume concentration distribution stored in the volume concentration distribution memory 410, and the calculation result is input to the control device 1000. On the other hand, the contents of the background / block brightness difference memory 710 are also input to the control device 1000. The controller 1000 receives the logarithmic mean particle size and the background / block luminance difference signal to calculate the increase / decrease of the coagulant injection amount, and inputs this signal to the coagulant injection pump 120.
Upon receiving this signal, the coagulant injection pump 120 operates the coagulant injection amount from the coagulant tank 110 to the rapid mixing basin 100. The startup brightness of the image processing startup timer 350 is 10
If once a minute, change the coagulant injection amount once every 10 minutes
Times.

The above is the basic configuration and operation of the present invention.

Next, the brightness distribution calculating means 5 which is the main point of the present invention
A series of image processing methods from 00 to the background / block luminance difference calculation means 700 will be described in detail with reference to the drawings.

FIG. 3 shows an image of the flocks recognized by the image pickup device 180 such as an underwater camera. Since the flock group is white, the brightness level is high, while the background is dark and the brightness level is low. Since FIG. 3 is a grayscale image, the boundary between the block and the background is not clear in reality, but only the outline of the block is shown for clarity. FIG. 4 shows the distribution of the brightness levels scanned by the line AA ′ on the screen of FIG. Here, the brightness level is shown in 128 steps,
It means that the brightness is low in the upper direction of the vertical axis, while the brightness is high in the lower direction, and the brightness is high because the aggregation block is white. That is, the downward valleys represent the flock. In the luminance distribution of FIG. 4, the binarization processing based on the luminance threshold value designated by the line BB is the binarization processing diagram of FIG.

As mentioned above, the rule of thumb of the maintenance manager of the water purification plant is that the sedimentation characteristics are good when the flock looks sluggish, but this is because the background brightness level in Fig. 4 is low.
This means that the brightness level of the flock is high. FIG. 6 shows this more specifically, and FIG. 6 (a)
Is a luminance distribution when the sedimentation property is good (when the adhesiveness is visible), and (b) is a luminance distribution when the sedimentation property is poor (when the adhesiveness is not visible). In each case, the binarized form is as shown in FIG. 5, and it is not possible to find the difference between the good sedimentation and the badness by simply obtaining the particle size and the particle size distribution. In the present invention, the brightness distribution of all pixels of the grayscale image shown in FIG. The results are shown in FIG. FIG. 7 (a) corresponds to FIG. 6 (a), and FIG. 7 (b) corresponds to FIG. 6 (b). With respect to the background / float luminance difference, the background / float luminance difference calculation means 700 obtains the luminance difference C from the difference between the peak values P _B and P _F of the number of pixels.

The control device 1000 receives the logarithmic mean particle size of the flocs calculated from the volume concentration distribution stored in the volume concentration distribution memory 410 and the luminance difference from the background / float luminance difference memory 710, and calculates these values as the first value. The coagulant injection amount is controlled so that it enters the Op region having excellent sedimentation characteristics shown in FIG. For example, when the logarithmic mean particle size of the flocks is too small or the difference in brightness between the background and the flocks is too large, the coagulant injection amount is increased.

〔The invention's effect〕

According to the present invention, the particle size distribution of the flocs and the brightness difference (contrast) between the background and the flocs are constantly measured quantitatively and the coagulant injection amount is manipulated. An excellent block can be formed with reliability. Therefore, the load on the settling basin and the filtration basin can be kept low, which in turn makes it possible to save energy, labor and reliability of the water treatment plant maintenance.

[Brief description of drawings]

FIGS. 1 and 2 are diagrams showing an embodiment of the present invention, FIGS. 3 to 7 are diagrams showing image processing contents of the present invention, and FIG. 8 is a diagram showing an optimum region for block formation control. Is. 150: Flock formation pond, 180: Imaging device, 30
0 ... Image recognition means, 400 ... Gray image information storage means, 410 ... Volume concentration distribution memory, 500 ... Luminance distribution calculation means, 700 ... Background / float luminance difference calculation means, 1000 ... Control device.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Shoji Watanabe 4026 Kuji Town, Hitachi City, Ibaraki Prefecture Hitachi Research Laboratory, Hitachi, Ltd. (56) References JP-A-61-21306 (JP, A) JP-A-61 -187907 (JP, A) JP-A-61-111111 (JP, A)

Claims (2)

[Claims] 1. An image capturing means for taking in image information of a flocks group in a flocks forming pond in which a flocculant is injected to form flocks of turbidity as an electric signal using a photoelectric conversion device, and the flocks based on the electric signal. Flock calculation means for image processing the group to calculate the floc particle size distribution and converting it to the flock volume concentration distribution, and for calculating the floc logarithmic mean particle size from the flock volume concentration distribution obtained for each of a plurality of images; Luminance computing means for computing the luminance of the block and the background based on the signal, luminance difference computing means for computing the luminance difference between the block and the background obtained by the luminance computing means, the logarithmic mean particle size of the floc and the background of the floc and the background Agglomeration so that the difference in brightness of the particles falls within a region with excellent floc sedimentation characteristics set in advance by the relationship between the logarithmic mean particle size of the particles and the difference in brightness. An injection control means for controlling the amount of agent injection, which is provided for monitoring and controlling the clean water block. 2. The brightness calculation means according to claim 1, wherein the brightness calculation means calculates the brightness of all pixels of the block group, the pixel and the representative brightness of the background are calculated from the pixel distribution with respect to this brightness, and the brightness difference calculation means A supervisory control device for clean water block, wherein the difference in the representative brightness is used as the brightness difference between the background and the block.