JP2867656B2 - Pump number control method by fuzzy inference - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION A. Industrial Field of the Invention The present invention relates to a pump number control method for instructing an increase or decrease in the number of pumps to be operated based on a forecast of inflow water into a well and a well water level in a sampling well of a water supply system or the like. More particularly, the present invention relates to a pump number control method using fuzzy inference.

B. Summary of the Invention The present invention relates to a method for controlling the number of pumps instructing an increase or decrease in the number of pumps to be operated based on the water level of a well or a reservoir in a water supply system and the amount of inflow water or the amount of water distribution. Predict water volume,
Judge whether the predicted inflow water amount or water distribution amount is within the optimal range to be set by judging from the discharge amount according to the current number of pumps operated.If the predicted inflow water amount or the predicted water distribution amount is within the optimum range, the current pump When operating with the number of operating units, the optimum time to be maintained between the lower control water level and the upper control water level is determined from the predicted inflow water amount or the predicted distribution water amount, and if the optimum time is sufficiently longer than the preset reference time, the pump is operated. The number of operating units is maintained as it is, and if the estimated inflow water amount or the predicted distribution amount is out of the optimum range and the unsuitable time is sufficiently long, the number of operating pumps is increased or decreased.If the optimum time or unsuitable time is not sufficiently long, , Fuzzy inference to judge the increase or decrease of the number of pumps operated or the status quo based on the current pump well or reservoir water level and the degree of water level rise or fall based on the forecast of inflow or distribution volume Therefore, the concept can be embodied without the need for seasoned operators, and the start / stop of the pump can quickly follow the rapid increase or decrease in the inflow water. It is intended to provide a technique that does not require a large level interval.

C. Prior Art The control method of the number of pumps currently employed in a pump well of a water supply system or the like measures the water level of the pump well and sets the number of operating pumps according to the water level.

It is necessary to increase or decrease the number of operating pumps according to changes in the amount of inflow water. For this reason, a change in the water amount is detected based on a change in the pump well level, and the number of pumps operated is changed to control the pump well level within a certain range.

FIG. 9 is a configuration diagram showing an example of a conventional pump number control circuit. In the figure, 91 is a pump well, 92 is a water level gauge, 93 is a converter, 94 is an alarm setting device, 95 is an operation sequence switching circuit, 96 is a pump operation circuit, and a 96 is a pump operation circuit. The signal is input to the alarm setting device 94 via the converter 93, and is converted into a pump operation signal, and the number of pumps operated is controlled by the operation sequence switching circuit 95. The alarm setting device 94 is, for example, a
Set as shown in Figure 10. In the figure, for example, N
The O.1 pump is started when the water level in the pump well is equal to or higher than L5 and is operated until the water level drops to L2 or lower. The inflow is larger than the capacity of the No. 1 pump, and the water level rises further
When L6 is reached, start the NO.2 pump. The NO.3 pump operates in the same manner.

D. Problems to be Solved by the Invention The above-mentioned conventional pump number control method has an advantage of being simple, but when the amount of inflow water fluctuates frequently, the number of starts / stops of the pump is increased, which is not preferable. Also, depending on the relationship between the amount of inflow water and the pump capacity, the frequency of start / stop of the pump increases even if the change in the amount of water is small. is there.

That is, the conventional pump number control method has the following problems.

(1) Although the prediction of the inflow water amount is accurate and an experienced operator can set the optimal number of operating units, it is not always possible to always set the number of experienced operators to set the number. No

(2) Since the number of pumps operated is determined based on the water level of the pump well, when the volume of the pump well is smaller than the discharge capacity of the pump, the start / stop of the pump follows the rapid increase and decrease of the inflow water volume. Can not.

(3) In unit control based on water level, when the number of pumps is large, it is necessary to increase the interval between HWL (high setting water level) and LWL (low setting water level), and the fluctuation of the pump well water level increases. In addition, if the interval is narrowed, the control cannot be performed unless the detection accuracy of the water level meter is very good, and even if the detection accuracy is cleared, the frequency of starting / stopping the pump is increased, which is not preferable.

The present invention has been made in view of such a problem, and the concept can be realized without a veteran operator, and the start / stop of the pump can be quickly performed in response to a sudden increase or decrease of the inflow water amount. It is an object of the present invention to provide a pump number control method which can follow the pump number and does not need to set the set level interval so much even if the number of pumps is large.

E. Means for Solving the Problems Means for solving the above problems in the present invention are:
In a pump number control method that commands the increase or decrease of the number of pumps to be operated based on the water level of a well or reservoir in a water supply system and the amount of inflow water or the amount of water distribution, the amount of inflow water into the well or the amount of water distribution from the reservoir is predicted, and the predicted inflow water Or, determine if the water distribution is within the optimal range to be set by judging from the discharge amount based on the current number of pumps operated, and operate with the current number of pumps operating if the predicted inflow water or the predicted water distribution is within the optimal range. At that time, the optimum time to be maintained between the lower control water level and the upper control water level is determined from the predicted inflow water amount or the predicted distribution water amount, and if the optimum time is sufficiently longer than the preset reference time, the pump operation number is If the predicted inflow or predicted water distribution is out of the optimum range and the unsuitable time is sufficiently long, increase or decrease the number of operating pumps.If the optimum time or the unsuitable time is not sufficiently long, The number of pumps operated or the current status is determined by fuzzy inference based on the current pump well or reservoir water level and the degree of rise or fall of the water level based on the forecast of inflow or water distribution. I do.

F. Action The present invention is a pump number control system that determines the number of pumps to be operated based on the prediction of the inflow water amount and the increase / decrease of the number of pumps based on fuzzy inference.

Fuzzy inference is a theory that has gained attention with the advancement of artificial intelligence (AI) technology. In AI, an automatic diagnosis system similar to humans can be configured in a certain field by storing the necessary knowledge base and rules of thinking judgment in a computer, but only accepting vague expressions like humans Can't do it, and you must give it a logically proper expression. Fuzzy control has been emphasized in order to compensate for this shortcoming and introduce ambiguity into computers. When expressing a height or size, even if a numerical value with a high possibility is adopted, Do not use the fixed numbers. Specifically, a membership function is used as a knowledge base, and a dedicated inference rule is provided.

In the present invention, the current water level and the water level fluctuation prediction are respectively set to, for example, five stages as a membership function, and the inference rules are based on their matrix configuration.
The following steps are performed.

(1) Predict the amount of inflow water.

(2) Determine whether the predicted inflow water amount is in the optimum range.

(3) If the estimated inflow water amount is in the optimum range, the number of pumps will maintain the current state if the optimum time is sufficiently long.

(4) If the estimated inflow water amount is out of the optimum range, the number of pumps is increased or decreased if the inappropriate time is sufficiently long.

(5) If the optimum time or the inappropriate time is not sufficiently long, it is determined whether the number of pumps is increased or decreased or the current status is maintained by fuzzy inference.

(6) If it is decided to increase or decrease the number of pumps, issue the command. The number is increased or decreased one by one.

G. Examples Hereinafter, examples of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a configuration diagram of one embodiment of the present invention. In the figure,
1 accumulates water level setting data and inflow water volume prediction data,
An industrial personal computer that performs a unit control arithmetic processing, 2 is a fuzzy inference controller in which rules and membership functions based on fuzzy inference are set, 3 is a man-machine interface, 4 is a process-side input / output device, and 5 is a link between them. LAN (Locai Area Network).

The industrial personal computer 1 accumulates data such as water level setting data and estimated inflow amount data and performs arithmetic processing for controlling the number of units, and has a LAN interface. The fuzzy inference controller 2 accumulates rules and membership functions and performs fuzzy inference. The man-machine interface 3 is a personal input / output device including a CRT, a keyboard, and a mouse, and performs various settings, system maintenance, and a result display. The process-side input / output device 4 receives the inflow prediction data and the current water level data, and outputs a one-unit increase command or a one-unit decrease command.

The above-described pump number control device can be incorporated in a monitoring control device or a local control device and can be shared, but in the present embodiment, as shown in FIG. n) 22 are controlled.

FIG. 3 is a flowchart showing an example of the number-of-pumps control method of the present invention using the above-described apparatus. In the figure,
The flow is started when the inflow amount prediction data is input to the industrial personal computer 1, and the personal computer 1 determines whether or not the predicted inflow water amount Q is in the optimum range. The optimum range is defined as follows: (k-1) q ≦ Q ≦ (k + 1) q, where q is the discharge water amount of each pump and k is the current number of operating units, as shown in FIG. To establish. Assuming that there are two types of pumps, large and small, and that there is only one pump with a small discharge amount (qs), as shown in FIG. 4 (b), the optimum range is as follows: a; B: optimal range for one small pump, c: optimal range for one large pump, d: optimal range for two large pumps, e: three large pumps Optimum range in the case of f, and Optimum range in the case of 4 to large pumps.

If the predicted inflow water amount is in the optimum range, it is next checked whether or not the optimum time TM is sufficiently long. The optimal time TM is defined as S, where S is the bottom area of the pump well, h is the current water level, LWL is the lower control water level, and HWL is the upper control water level. Is the longest time that is established, that is, if the operation is performed with the current number of vehicles and the predicted inflow water amount Q is correct, it means the time during which the water level is maintained between LWL and HWL. The reference time T ₀ is set in advance, and if TM ≧ T ₀ , the optimal time is determined to be sufficiently long, and TM
If <T ₀ , the optimal time is assumed to be short. If the optimal time is long enough, the number of pumps will maintain the current status.

An unsuitable time and its reference time are set in the same way, and if the estimated inflow water amount Q is out of the optimum range, the number of pumps is increased or decreased if the unsuitable time is sufficiently long.

If the optimum time or the inappropriate time is not long, it is determined whether the number of pumps is increased or decreased or the current status is maintained by fuzzy inference.

As already mentioned, fuzzy inference does not specify numerical values. In this embodiment, as a membership function,
NB, NS, M, PS, PB
, And inference rules are set between them as shown in the matrix in Table 1.

The above table shows the IF to THEN rules in a matrix. For example, IF to the current water level is PS (slightly closer to the management upper limit).
Then, if the water level fluctuation prediction is PB (the water level rises considerably), then THEN-the number of pumps is PB (need to increase), indicating 5 × 5 = 25 rules.

Hereinafter, the fuzzy inference in the present invention will be described in more detail.

FIG. 5 is an explanatory diagram of the membership function. As shown in FIG. 3A, HH (upper limit), H (upper limit of management), L (lower limit of control), and LL (lower limit) are set for the water level h, and a range from H to L is defined as a water level control range. On the other hand, the membership function of the current water level, as shown in FIG.

NS; slightly closer to the lower control limit.

M; middle.

PS; slightly closer to the management limit.

PB; Very close to management limit.

Also, in the membership function for water level fluctuation prediction, as shown in FIG.

NS; water level falls slightly.

M; Water level does not change.

PS; water level rises slightly.

PB; water level rises considerably.

On the other hand, the membership function for increasing or decreasing the number of pumps, as shown in FIG.

NS; The number of pumps needs to decrease slightly.

M; The number of pumps may be maintained as it is.

PS; The number of pumps needs to increase slightly.

PB; The number of pumps needs to be increased.

It is.

FIG. 6 and FIG. 7 are explanatory diagrams of inference rules. FIG. 6 (a) is a diagram for studying the predicted inflow water amount, in which a straight line indicates the pump discharge amount q and a curve indicates the predicted inflow water amount.
The upper dotted line shows the upper management limit of the current number, and the lower dotted line shows the lower management limit of the current number. Assuming that the current time is 0, a period in which the predicted inflow amount exceeds the management upper limit between t1 and t2 is predicted, so that it is necessary to increase the number at t1 and decrease the number at t2. Time T is the third
As shown in the flowchart of the figure, this is the time from when a start or stop command is issued until the operation is completed, and the time t2 is set earlier in consideration of this. FIG. 6 (b)
Is a time characteristic diagram of the water level corresponding to the above.

FIG. 7 shows fuzzy inference for the predicted inflow water amount and the water level setting as shown in FIG. At t = t1, if the water level is h 'between LWL and HWL in the current water level diagram shown in FIG. 7 (a), the necessity of changing the number of pumps (0 to 1) is represented by a triangle having PS as a vertex. And a line of intersection with a triangle having PB as a vertex. In addition, in the water level fluctuation prediction diagram shown in FIG. 7B, if it is h ″ between fluctuations −h to + h,
Similarly, an intersection line with a triangle having M as a vertex and an intersection line with a triangle having PS as a vertex are obtained. According to the rule matrix, the necessity of changing the number of pumps (−1 to +1) is shown in FIG. As shown by the hatched area in (c), it is shown on the plus side. If the inference result is a necessity of 0.75 or more, it is determined that one pump is added, and if the necessity is -0.75 or less, it is determined that one pump is reduced.

FIGS. 7 (d), (e) and (f) show fuzzy inference at t = t2, and the inference result is shown on the minus side as shown in FIG. You.

As described above, when it is determined that the number of pumps is to be increased or decreased or the current status is to be maintained, the flow branches respectively, and when increasing or decreasing the number of pumps, a command to increase or decrease the number of pumps by one is issued.

In this embodiment, the control of the number of pumps for eliminating inflow water is shown. However, the control method of the present invention is also applicable to the control of the number of pumps 83 for supplying water from the water purification reservoir 81 to the distribution reservoir 82, as shown in FIG. Applicable. In that case, HWL and LWL should be replaced, and inflow prediction should be replaced with water distribution prediction. The control method of the present invention is based on the premise that some kind of water amount prediction is performed and that the prediction is somewhat accurate. However, in the case of rainwater, the prediction method is a rainfall runoff analysis model or a pump well water level. There are predictions based on changes, etc., and in the case of water distribution, there is a demand prediction, and the effect increases by combining these predictions. Further, the function of the present invention can be incorporated in a monitoring control device or a local control device.

 This embodiment has the following advantages.

(1) Since the know-how of the operation of the pump by the experienced operator is expressed by the rule and the membership function, it is possible to set the number of pumps close to the concept of a human.

(2) The start / stop of the pump can be operated quickly (with the shortest delay) for a sudden increase or decrease of the inflow water amount.

(3) Even if the number of pumps is large, the number of pumps can be controlled without taking a large space between HWL and LWL.

(4) Both a fixed speed pump and a variable speed pump can be applied, and it is possible to control even if the capacity of a plurality of pumps is the same capacity or a combination of different sizes.

(5) Fuzzy inference rules and membership functions are CRT
And can be easily added and changed using the keyboard and mouse.
Various settings such as T ₀ and S can also be performed by CRT operation.

(6) The features of the pump well volume, pump capacity and number, and inflow pattern can be easily expressed by fuzzy inference rules and membership functions, and can be packaged as general-purpose software that enables flexible system maintenance.

H. Effects of the Invention As described above, according to the present invention, it is possible to realize the concept without the need for a seasoned operator, and to quickly start / stop the pump when the amount of inflow water increases or decreases rapidly. It is possible to provide a system for controlling the number of pumps, which can be made to follow, and does not need to have a wide set level interval even if the number of pumps is large.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention, FIG. 2 is a block diagram of an application example of the present invention, FIG. 3 is a flowchart of one embodiment of the present invention, and FIG. , FIG. 5 is an explanatory diagram of a membership function, FIGS. 6 and 7 are explanatory diagrams of inference rules, FIG. 8 is a configuration diagram of another application example of the present invention, and FIG. 9 is a configuration diagram of a conventional example. FIG. 10 is an explanatory diagram of a pump operation water level. 1 ... Industrial personal computer, 2 ... Fuzzy inference controller, 3 ... Man-machine interface, 4 ... Process side input / output device, 5 ... LAN, 21 ... Pump well, 22 ...
... Pump, 81 ... Pure water reservoir, 82 ... Distribution reservoir, 83 ... Pump, 91 ... Pump well, 92 ... Water gauge, 93 ... Transducer, 94
…… Alarm setting device, 95 …… Operation sequence switching circuit, 96 …… Pump operation circuit.

Continuation of front page (56) References JP-A-1-246601 (JP, A) JP-A-2-131181 (JP, A) JP-A-59-90787 (JP, A) JP-A-59-3186 (JP) , A) JP-A-58-82317 (JP, A) Toshiro Terano, two others, "Introduction to Applied Fuzzy Systems", 1st edition, Ohm Co., Ltd., May 25, 1989, p. 97-110 (Section 3.5 “Rainwater Pump Management”) (58) Field surveyed (Int. Cl. ⁶ , DB name) G05B 13/00-13/04 G05B 11/18 F04B 49/06 321 JICST file ( JOIS)

Claims (1)

(57) [Claims] A pump number control method for instructing an increase or decrease in the number of pumps to be operated based on the water level of a well or a distribution reservoir and an inflow water amount or a distribution water amount in a water supply system, wherein the amount of inflow water into the well or the distribution water amount from the distribution reservoir is predicted. It is determined whether or not the predicted inflow water amount or the distribution water amount is within the optimal range to be set by judging from the discharge amount based on the current pump operation number. When operating with the number of operating pumps, determine the optimum time to be maintained between the lower control water level and the upper control water level from the predicted inflow water amount or the predicted water distribution amount, and if the optimum time is sufficiently longer than the preset reference time. The number of operating pumps is maintained at the current level, and if the estimated inflow water amount or the estimated distribution water amount is out of the optimum range and the unsuitable time is sufficiently long, the number of operating pumps is increased or decreased, and the optimum time or unsuitable time is sufficiently long. If not, it is characterized by the fuzzy inference of the increase or decrease of the number of pumps operated or the maintenance of the current status based on the current pump well or reservoir water level and the degree of rise or fall of the water level based on the forecast of inflow or distribution volume. Pump number control method by fuzzy inference.