JP2018185678A - Operation planning device, operation control system, and operation planning method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technique capable of devising an operation plan timely reflected with individual plant situations and operation know-how.SOLUTION: An operation planning device 101 includes: an operation range extraction unit 111 which extracts an operation range distribution related to a prescribed event from operation achievement data related to a plant and defines a basic index indicating a deviation degree from the operation range per prescribed event; a problem configuration learning unit 112 which generates an evaluation index by combining basic indexes of the prescribed event in an operation plan, devises the operation plan about past prescribed periods by solving an optimization problem in which constraint conditions related to the plant configuration and the evaluation index are used as an object function, calculates a matching degree between the operation plan and operation achievement data related to the past prescribed periods, and searches for an evaluation index configuration in which the matching degree becomes the maximum; and an operation plan device unit 114 which devises the operation plan by solving an optimization problem in which the searched evaluation index is used as the object function on the basis of a prescribed demand prediction value related to the plant.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an operation plan planning device, an operation control system, and an operation plan planning method, and more specifically, a technology that makes it possible to plan an operation plan that appropriately reflects individual plant conditions and operation know-how. About.

  When operating various plants such as water and sewage facilities, formulate an operation plan that optimizes the quality and efficiency of operation, taking into account various conditions such as the specifications of the facility, the supply and demand relationship of products, operating resources, etc. There is a need.

  As a conventional technique related to such an operation plan, for example, in a plant operation control apparatus that performs control for plant equipment in a plant based on an operation plan that has been created in advance, process data input from the plant via a process controller is used. A process data storage means for storing, a past performance value stored in the process data storage means, and a demand prediction section for predicting and calculating demand based on predetermined information input from a human interface; and the demand prediction section IA optimal plan calculating means for calculating an optimal operation plan using an immune algorithm based on the predicted demand calculation value, and when the IA optimal plan calculating means performs calculation using the immune algorithm , Cells for storing memory cells and suppressor cells And 憶 means, such as a plant operation control apparatus characterized by comprising a (see Patent Document 1) has been proposed.

JP 2003-228402 A

  In the prior art, it is possible to obtain a semi-optimal operation plan at high speed by using the actual value as the initial solution for optimization. However, it does not assume a plant with a wide range of facilities such as distribution reservoirs and pumps as the target, and the operation plan of skilled operators who should be latent in the operation results is appropriately set in the operation plan. There is no reflection. That is, from the viewpoint of optimizing the operation plan, there remains a problem that the events to be reflected in the operation plan are not sufficiently reflected. These issues are particularly evident in water supply businesses that have problems such as difficulties in maintaining the business due to future population decline, and aging of skilled workers and uninherited operation know-how due to retirement.

  Therefore, an object of the present invention is to provide a technique that enables the creation of an operation plan that appropriately reflects the individual plant situation and operation know-how.

The operation plan planning apparatus of the present invention that solves the above problems is a device for planning a plant operation plan, which extracts a distribution of an operation range related to a predetermined event from operation result data related to the plant, and deviates from the operation range. By combining the operation range extraction unit that defines the basic index for each predetermined event and the basic index of the predetermined event, an evaluation index in the operation plan is generated, and the constraint condition regarding the configuration of the plant and the evaluation index By solving the optimization problem with the objective function as an objective function, an operation plan is drafted for a predetermined period in the past, the degree of coincidence between the operation plan and the operation result data for the period is calculated, and the degree of coincidence is maximized. A problem configuration learning unit that searches for a configuration of the evaluation index, and an objective function that determines the searched evaluation index based on a predetermined demand prediction value related to the plant Characterized in that it comprises the operation planning unit, the to design operation plan by solving the optimization problem.

  Further, the operation control system of the present invention is an operation control system for controlling a plant based on an operation plan, which collects measurement values by the sensors of the plant and performs a control instruction to a sub-controller that controls the plant. And an operation plan planning device for planning an operation plan as a control target of the monitoring control device, from an operation result data as a measurement value obtained through the monitoring control device, and an operation range related to a predetermined event And generating an evaluation index in the operation plan by combining a basic index of the predetermined event and an operating range extraction unit that determines a basic index indicating the degree of deviation from the operating range for each predetermined event. By solving an optimization problem with the objective function of the constraint condition regarding the plant configuration and the evaluation index, A problem configuration learning unit that formulates an operation plan, calculates a degree of coincidence between the operation plan and the operation result data relating to the period, and searches for a configuration of the evaluation index that maximizes the degree of coincidence; An operation plan drafting device comprising an operation plan drafting unit for drafting an operation plan by solving an optimization problem using the searched evaluation index as an objective function based on a predetermined demand forecast value relating to To do.

  Further, according to the operation plan planning method of the present invention, the information processing device that plans the plant operation plan extracts the distribution of the operation range related to the predetermined event from the operation result data related to the plant, and indicates the deviation from the operation range. By combining the operation range extraction process for defining the basic index for each predetermined event and the basic index of the predetermined event, an evaluation index in the operation plan is generated, and the constraint condition regarding the configuration of the plant and the evaluation index are used as the objective function By solving the optimization problem, an operation plan is drafted for a predetermined period in the past, the degree of coincidence between the operation plan and the operation result data for the period is calculated, and the evaluation that maximizes the degree of coincidence Based on the problem configuration learning process for searching for the configuration of the index and a predetermined demand prediction value for the plant, the searched evaluation index is used as an objective function. And characterized in that the execution and operation planning process for planning the operation plan by solving the optimization problem, the.

  According to the present invention, it is possible to make an operation plan that appropriately reflects individual plant conditions and operation know-how.

1 is a diagram illustrating a logical configuration example of an operation control system in Embodiment 1. FIG. 1 is a diagram illustrating an example of a hardware configuration of an operation plan planning apparatus in Embodiment 1. FIG. It is a figure which shows the structural example of the water supply facilities group used as the operation control object in Example 1. FIG. It is a figure which shows the example of the operation plan in Example 1. FIG. It is a figure which shows the example of the operation plan in Example 1. FIG. It is a figure which shows the example of the distribution of the driving performance in Example 1, and a basic index value. It is a figure which shows the example of a flow of the operation plan planning method in Example 1. FIG. 6 is a diagram illustrating an example of a problem configuration display window in Embodiment 1. FIG. It is a figure which shows the example of the operation plan display window in Example 1. FIG. It is a figure which shows the logical structural example of the operation control system in Example 2. FIG. It is a figure which shows the example of a flow of the operation plan planning method in Example 2. FIG. It is a figure which shows the example of the problem structure display window classified by operation conditions in Example 2. FIG.

--- Example 1 ---
Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a diagram illustrating a logical configuration example of an operation control system 100 including an operation plan planning apparatus 101 according to the first embodiment. The operation control system 100 in the present embodiment is a system that controls each device in a water plant based on an operation plan previously planned by the operation plan planning device 101 and manages a delivery / distribution process of the water plant. In this embodiment, the target of the operation plan and the operation control is the water plant, but it is not limited to this.

  As illustrated in FIG. 1, the operation control system 100 according to the present exemplary embodiment includes an operation plan planning device 101, a monitoring control device 102, a sensor 103, and a sub-controller 104.

  Among them, the operation plan planning device 101 includes an operation range extraction unit 111, a problem configuration learning unit 112, a demand prediction unit 113, an operation plan planning unit 114, an operation result data storage unit 121, a data collection unit 151, a plan output unit 152, and , A configuration / plan display unit 153.

  The operation range extraction unit 111 described above acquires past operation result data related to the water plant from the operation result data storage unit 121, extracts the distribution of the operation range related to the predetermined event from the operation result data, and the operation range. A basic index indicating the degree of deviation from the above is determined for each predetermined event described above. In addition, the driving range extraction unit 111 transmits information on the basic index determined as described above to the problem configuration learning unit 112 and the configuration / plan display unit 153. Details of the processing of the operation range extraction unit 111 will be described later together with the description of FIG.

  In addition, the problem configuration learning unit 112 obtains past operation result data related to the water plant from the operation result data storage unit 121 and combines the basic indicators obtained from the operation range extraction unit 111 described above. By generating a plan evaluation index and solving an optimization problem with the objective function of the constraint condition regarding the configuration of the water plant and the above-mentioned evaluation index, an operation plan is drafted for a predetermined period in the past. The degree of coincidence with the operation result data relating to the period is calculated, and the configuration of the evaluation index that maximizes the degree of coincidence is searched. In addition, the problem configuration learning unit 112 transmits the configuration of the evaluation index with the maximum degree of coincidence thus obtained as a problem configuration to the operation plan planning unit 114 and the configuration / plan display unit 153. Details of the processing of the problem configuration learning unit 112 will be described later together with the description of FIG.

  In addition, the demand prediction unit 113 acquires past operation result data related to the water plant from the operation result data storage unit 121, calculates a predicted value of demand based on the operation result data, and calculates the predicted value of this demand. Is transmitted to the operation planning unit 114. Here, the predicted value of the demand is a predicted value of the water distribution amount for each water distribution area in the water plant at every predetermined time step (for example, around 30 minutes) in a predetermined future period (for example, 24 hours) from the prediction time point. Refers to time series. A known time series prediction method is adopted as a calculation method for such demand prediction.

  In addition, the operation planning unit 114 acquires the latest operation result data regarding the water plant from the operation result data storage unit 121, acquires the above-described problem configuration obtained from the problem configuration learning unit 112, and demand prediction The demand forecast value is acquired from the unit 113, the constraint condition to be satisfied by the operation plan is calculated from the demand forecast value and the latest operation record data, and the constraint condition and the problem configuration learning unit 112 described above are calculated. An operational plan is created by solving an optimization problem using an evaluation index as an objective function. Further, the operation plan drafting unit 114 transmits the drafted operation plan to the plan output unit 152 and the configuration / plan display unit 153. Details of the processing of the operation planning unit 114 will be described later together with the description of FIG. Details of the operation plan will be described later with reference to FIG.

In addition, the operation result data storage unit 121 acquires the operation result data of the delivery / distribution process subject to the operation plan in the water plant from the data collection unit 151, accumulates the operation result data, and receives the operation result when requested. The data is transmitted to the operation range extraction unit 111, the problem configuration learning unit 112, the demand prediction unit 113, and the operation plan planning unit 114.

  In addition, operation performance data refers to sensor measurement values of a water distribution process in a water plant, data on failure states of devices and sensors, and the like. Further, the operation result data may include related data managed by the monitoring control apparatus 102, that is, alarm information related to a change in the failure state of the device or sensor.

  In addition, the data collection unit 151 receives operation result data of the delivery / distribution process of the operation plan target in the water plant from the monitoring control device 102 and transmits the operation result data to the operation result data storage unit 121. It is assumed that the data collection unit 151 and the monitoring control apparatus 102 are connected to be communicable via an appropriate network.

  The plan output unit 152 acquires an operation plan from the operation plan planning unit 114 and transmits the operation plan to the monitoring control apparatus 102 as a target value for operation control.

  In addition, the configuration / plan display unit 153 acquires the operation range distribution and basic index information from the operation range extraction unit 111, acquires the problem configuration from the problem configuration learning unit 112, and the operation plan planning unit 114. The operation plan is acquired, and the acquired information is displayed in a window on a display for an appropriate operator such as the display unit 206 described later. Display examples of the configuration / plan display unit 153 will be described later with reference to FIGS. 7 and 8.

  In addition, the monitoring control device 102 collects measurement values from the sensor 103 that measures the state of the delivery / distribution process in the operation-targeted water plant, and the measured value and the failure state of the sensor 103 and the equipment in the water plant. This is a device that transmits information and alarm information related to changes in the failure state of the sensor 103 or the device to the operation planning device 101. The monitoring control device 102 also receives an operation plan from the operation plan planning device 101 as a target value for operation control, and controls the state of the delivery / distribution process in the water plant subject to operation control based on the operation plan. A control target value is transmitted to the controller 104.

  The operation planning device 101 may be realized as a program on a data center installed at a remote location from the monitoring control device 102. Further, the operation planning apparatus 101 may be realized by being divided into a program on a remote data center and a terminal placed on the monitoring control apparatus 102 side. For example, the driving range extraction unit 111 and the problem configuration learning unit 112 can be implemented by a program on a data center, and the above-described terminals can be implemented except for the driving range extraction unit 111 and the problem configuration learning unit 112.

  Next, the hardware configuration of the operation plan planning apparatus 101 in this embodiment will be described. FIG. 2 is a diagram illustrating an example of a hardware configuration of the operation planning apparatus according to the first embodiment.

  As illustrated in FIG. 2, the operation planning apparatus 101 in this embodiment includes a CPU 201, a memory 202, a media input / output unit 203, an input unit 205, a communication control unit 204, a display unit 206, a peripheral device IF unit 207, and , And a bus 210.

  Among these, the CPU 201 is an arithmetic device that executes the program 2021 on the memory 202. The memory 202 is storage means for temporarily storing the above-described program 2021 and tables. The media input / output unit 203 holds the above-described program 2021 and table read from a predetermined storage medium, and stores them in the memory 202 as appropriate.

  The input unit 205 is a device that provides a user with an input interface such as a keyboard and a mouse. The communication control unit 204 is a network interface or the like that is connected to the network 220 and executes communication with an external device such as the monitoring control device 102.

  The display unit 206 is a device that is a display destination of information by the configuration / plan display unit 153 in FIG. 1, and is specifically a display. The peripheral device IF unit 207 is an interface such as a printer.

  The bus 210 interconnects the CPU 201, the memory 202, the media input / output unit 203, the input unit 205, the communication control unit 204, the display unit 206, and the peripheral device IF unit 207.

  As is clear from the comparison between FIG. 1 and FIG. 2, the operation planning apparatus 101 in FIG. 1 is realized by the CPU 201 executing the program 2021. The operation planning device 101 and the monitoring control device 102 may be realized as different programs on the same hardware.

  FIG. 3 is a diagram illustrating an equipment configuration of a delivery / distribution system in a water plant that is subject to operation control in the first embodiment. With reference to FIG. 3, the structure of the water supply / distribution process of the water supply which is the water supply plant which the operation plan planning apparatus 101 monitors is demonstrated.

  In the equipment configuration illustrated in FIG. 3, the water is distributed from the distribution reservoirs 301 and 302 to the distribution area 341 by multi-point injection by the distribution pump facilities 321 and 322. In the figure, only a representative water pipe, such as a water pipe 354, is shown.

  Moreover, in the above-mentioned water distribution area 341, there are terminal stations 361 and 361, which measure water pressure and water quality. In addition, the natural water is distributed from the distribution reservoir 303 to the distribution area 342 through the pressure reducing valve 334. On the other hand, the inflow to the distribution reservoir 301 includes water received from the water supplier through the water pipe 351 and purified water obtained by treating the raw water from the intake pump facility 324 at the water purification plant 311. A part of the water in the distribution reservoir 301 is supplied to the distribution reservoir 303 via the water supply pump facility 323 and the water supply pipe 353. In addition, the inflow to the distributing reservoir 322 is received from the water supply company through the water pipe 352. In addition, the flow volume adjustment valve | bulb 331,332,333 is arrange | positioned in the inflow part from the water pipe of each distribution reservoir.

  In this example, the sensor 103 for collecting the measurement values by the monitoring control device 102 includes a water level meter in each reservoir, each pump facility, a flow meter in the valve, a water pressure meter in each pump facility and the terminal station, and a pump in each pump facility. The operation number, the number of revolutions, the measuring device of power consumption, the measuring device of the opening degree of each valve, etc. correspond.

  In this example, the sub-controller 104 to which the monitoring and control device 102 instructs the control corresponds to a controller that controls the start / stop of each pump, the rotation speed adjusting device, the opening adjustment of each valve, and the like.

  Next, an operation plan example according to the first embodiment will be described with reference to FIGS. 4A and 4B. Here, the operation plan refers to a plan for transition of control target or process state in a predetermined future period (for example, 24 hours) from the time of the operation plan. Specific examples of such control targets or process states include the water level in the distribution reservoir, the amount of water delivered and delivered by each flow meter, the number of pumps operated at each pump facility, the number of revolutions, and the opening of each valve.

4A shows a transition 400 of the water level of the reservoir A as an example of the operation plan, and FIG. 4B shows a transition 410 of the number of pumps operated in the water pump facility S as an example of the operation plan.

  The transition in such an operation plan is expressed by a time series of values for each time step (hereinafter, 30 minutes as an example). At this time, in order to express the process that changes every moment as a value every 30 minutes, a value representative of the time step is determined by an appropriate aggregation method. For example, the water level takes values at 0 and 30 minutes per hour, and the water intake / water supply / distribution amount takes an average value of 30 minutes each. If there is no surplus in the capacity of the reservoir in the target, the number of pumps to be operated needs to be switched not only at 0 and 30 minutes per hour, but also at any time.

  Next, with reference to FIG. 5, an example of the distribution of the operation range and the basic index extracted by the operation range extraction unit 111 of the present embodiment will be described.

  The operation range extraction unit 111 in this embodiment acquires operation result data from the operation result data storage unit 121, and performs processing in each step of operation range extraction, basic index definition, and result transmission based on this.

  Among these, in the operation result data acquisition step, the operation range extraction unit 111 acquires the operation result data from the operation result data storage unit 121.

  In the operation range extraction step, the operation range extraction unit 111 calculates a typical operation range of the operation result data for each target item. The item to be processed by the operation range extraction unit 111, that is, the target item is a measurement value of a process state set in advance. For example, the operation range extraction unit 111 targets events such as the amount of water intake, the amount of water delivered, the amount of water distributed, the water level of the reservoir, the number of operation switching operations of various pumps and valves, the interval, the time zone, and the operation time.

  In addition, the driving range extraction unit 111 can extract a plurality of driving range distributions and define a basic index with respect to the same measurement value by using a plurality of time sections and aggregation methods. The operation range extraction unit 111 can extract, for example, an operation range of a 30-minute average value and an operation range of an integrated value per day for water intake, and can define different basic indexes for each.

  In addition, the operation range extraction unit 111 may extract the distribution of the operation range for the value obtained by processing the measurement value, and determine the basic index. For example, the operation range extraction unit 111 includes the amount of change from the previous time of the water supply amount, the power consumption required for the operation of the delivery / distribution pump, and the 30-minute average value (power demand value) of the power in each facility based on the power consumption. In addition, an operation plan robustness index against an error in demand prediction can be processed.

  As for the robustness of the operation plan against demand forecast errors, for example, simulation of the transition of the reservoir water level when the operation operation is continued according to the operation plan in a situation where the demand is continuously greater or less than the forecast. And the maximum magnitude | size etc. from which a reservoir water level deviates from an operation range within predetermined time can be used.

  For example, the operation range extraction unit 111 sets the number of pumps to be switched per day in a certain distribution pump facility as a target item, calculates a histogram of the item, and calculates a range including most of the actual data.

  The (a) operation performance distribution graph 500 shown in FIG. 5 is a histogram of the number of times of switching the number of pumps per day in a certain distribution pump facility, calculated by the operation range extraction unit 111. The operation range extraction unit 111 calculates the range of the item by calculating the upper limit value of the number of switching times in which the relative frequency is equal to or greater than a predetermined value. The upper limit value of the range calculated by the operation range extraction unit 111 is indicated by a broken line in the figure.

  As a range calculation method by the operation range extraction unit 111, a method of calculating a predetermined percentile value, a method of calculating a value that is a predetermined multiple of the standard deviation from the average value, or the like is used. Further, not only the upper limit value but also the lower limit value is calculated according to the target item. If necessary, a plurality of upper limit values and lower limit values may be calculated for one target item.

  In addition, the operation range extraction unit 111 determines a basic index calculation method for quantifying the degree of deviation from the above-described operation range in the basic index definition step. For example, the operation range extraction unit 111 determines the following calculation method as a basic index using the operation range (the upper limit value of the number of switching times) regarding the number of pump switching times described above.

(Basic index value) = max [(number of times of switching) − (upper limit number of times of switching), 0]
Here, max is a maximum value calculation.

  A (b) basic index value graph 510 in FIG. 5 is a basic index value of the number of times of switching the number of pumps per day in a certain distribution pump facility, which is determined by the operation range extraction unit 111 as described above.

  In addition, the driving range extraction unit 111 transmits the list of basic indexes and the calculation method thereof to the problem configuration learning unit 112 and ends the process.

  The operation planning apparatus 101 can extract the operation tendency of an (experienced) operator for each operation range extraction item based on the past operation result data by the processing of the operation range extraction unit 111 described above. Note that the operation plan planning apparatus 101 can formulate an operation plan capable of continuing stable operation even if an error occurs in the demand prediction by using the robustness of the operation plan with respect to the error in the demand prediction as a basic index.

  Hereinafter, actual procedures of the operation planning method according to the present embodiment will be described with reference to the drawings. Various operations corresponding to the operation planning method described below are realized by a program that the operation planning device 101 reads into the memory or the like and executes. And this program is comprised from the code | cord | chord for performing the various operation | movement demonstrated below.

  FIG. 6 is a flowchart illustrating an operation plan planning method according to the first embodiment, and more specifically, a flowchart illustrating a processing flow of the problem configuration learning unit 112. Hereinafter, the processing of the problem configuration learning unit 112 will be described with reference to the flow of FIG.

  In this case, first, the problem configuration learning unit 112 starts processing (step 601). This start is executed by receiving a predetermined instruction from a predetermined device via the input unit 205 or the communication control unit 204, or detecting the arrival of a predetermined time.

  Subsequently, the problem configuration learning unit 112 acquires operation result data from the operation result data storage unit 121, and acquires a list of basic indexes and their calculation methods from the operation range extraction unit 111 (step 602).

  In addition, the problem configuration learning unit 112 generates an operation plan evaluation index by combining the basic indexes obtained in step 602 described above (step 603). The problem configuration learning unit 112 in this step generates at least one evaluation index by randomly combining basic indices by, for example, linear calculation and maximum value calculation.

Assuming that the vector representing the operation plan is X, the plant state up to the time of planning the operation plan is s, and the basic indicator or an indicator combining the basic indicators is h_i (where _ is an abbreviation for subscript), linear Examples of indicators that can be configured with a combination of calculation and maximum value calculation

がある。ここで、ｗ＿ｉ、ｋ＿ｉは任意の定数である。

There is. Here, w_i and k_i are arbitrary constants.

  Next, the problem configuration learning unit 112 drafts an operation plan for a predetermined period in the past by solving an optimization problem using the constraint conditions regarding the configuration in the water plant and the evaluation index obtained in step 603 as an objective function. (Step 604).

  In this case, the problem configuration learning unit 112 sets the evaluation index obtained in step 603 as f, and sets a plurality of constraint conditions related to the configuration of the water plant as g_j (j = 1, 2,..., N). For a certain past period p, if the vector representing the state of the plant that affects the period is s_p, the optimization problem

を解いて最適な運用計画案Ｘ＊＿ｐを立案する。

To formulate an optimal operation plan X * _p.

  Specifically, in the above-described state s_p, the demand amount (actual value or predicted value) in the period p, the reservoir water level and the number of pumps operated immediately before entering the period, and the amount of water intake before entering the period Contains values etc.

  In addition, examples of constraints related to water plant configuration include ensuring the amount of water distribution that satisfies water demand, that the amount of water delivered and distributed is within the design range of each facility, and that the water level of the distribution pond is within the design range of the facility In other words, the relationship between the water supply and distribution and the distribution reservoir water level may be satisfied. As a method for formulating these constraint conditions, a publicly-known technique relating to the planning of a delivery / distribution operation plan can be used.

  As a method for obtaining an operation plan as an optimal solution of the optimization problem, for example, a known technique such as a genetic algorithm or mathematical programming can be used.

  The problem configuration learning unit 112 prepares a plurality of periods p for evaluation in the next step, and drafts an operation plan for each period.

  Subsequently, the problem configuration learning unit 112 calculates the degree of coincidence between the operation plan drafted in step 604 described above and the operation result data in the period, and searches for the configuration of the evaluation index that maximizes the degree of coincidence. (Step 605).

  The problem configuration learning unit 112 in this step evaluates an operation plan drafted using the evaluation index f as an objective function, for example, based on the degree of coincidence given by the following equation.

ここで、Ｅ［ｆ］は評価値、ａ＿ｐは期間ｐにおける運用計画案Ｘ＊＿ｐに対応した実績値，［・］＿ｍはベクトルの第ｍ成分、ｌ＿ｍ，αは定数である。

Here, E [f] is the evaluation value, a_p is the actual value corresponding to the operation plan X * _p in the period p, [•] _m is the m-th component of the vector, and l_m and α are constants.

  The above equation of coincidence is calculated based on the amount of water intake, the amount of water delivered, the amount of water delivered, the water level of the reservoir, the operation switching time of the pumps and valves by the method of setting the constant l_m.

  Subsequently, the problem configuration learning unit 112 obtains an evaluation index f having a sufficiently high matching score based on the matching score E [f] calculated for the evaluation index f configured so far, and performs the problem configuration processing. It is determined whether or not to end (step 606).

  As a result of the determination described above, when the process ends (step 606: end), the problem configuration learning unit 112 causes the process to transition to step 607. On the other hand, when continuing as a result of the above-mentioned determination (Step 606: Continuation), the problem configuration learning unit 112 returns the processing to Step 603.

  For a specific method for determining the end of processing in step 606 and a method for constructing an evaluation index in step 603, a known technique such as genetic programming can be used.

  Next, the problem configuration learning unit 112 selects the evaluation index that maximizes the degree of coincidence among the evaluation indexes that have been configured so far, and the evaluation index selected as the objective function and the above-described constraints on the configuration of the water plant The set is determined as a problem configuration, and the problem configuration is transmitted to the operation planning unit 114 and the configuration / plan display unit 153 (step 607), and the flow is terminated (step 608).

  The operation planning apparatus 101 determines the overall request and determination of the operator regarding each operation range extraction item that may have a trade-off relationship with each other by the processing of the problem configuration learning unit 112 described above. As an addition, it can be extracted based on past driving performance data.

  Further, the operation plan planning unit 114 performs an operation plan planning process based on the problem configuration received from the problem configuration learning unit 112 with a predetermined future period from the plan planning time as a target for the plan planning target period. The formulation of the optimization problem and the algorithm for solving the optimization problem are the same as the processing in step 604 of the problem configuration learning unit 112.

  In addition, the operation plan planning apparatus 101 can operate the operation range extraction unit 111 regardless of the combination of the planned initial water level of the reservoir, the planned initial operation number of pump facilities, and the like by the processing of the above-described operation plan planning unit 114. In addition, an operation plan reflecting the past requests and judgments of (experienced) operators extracted by the problem configuration learning unit 112 can be created.

  Note that the driving range extraction unit 111 and the problem configuration learning unit 112 perform the processing at a predetermined cycle, for example, every several months. In addition, the operation planning unit 114 performs the processing in response to an instruction from the operator according to the degree of dissociation from the plant operating state in principle with a period of 24 hours or 12 hours. In the period between the processing periods of the operation range extraction unit 111 and the problem configuration learning unit 112, the operation plan planning unit 114 plans an operation plan using the problem configuration calculated by the problem configuration learning unit 112 last.

  FIG. 7 is a diagram illustrating an example of the problem configuration display window 701. With reference to FIG. 7, the configuration of the problem configuration displayed by the configuration / plan display unit 153 on the display unit 206 and the like will be described.

  A problem configuration display window 701 displayed on the display unit 206 such as a display by the configuration / plan display unit 153 has an objective function display panel 711. The configuration / plan display unit 153 displays the configuration of the objective function (evaluation index) among the problem configurations obtained from the problem configuration learning unit 112 on the objective function display panel 711.

  This objective function display panel 711 shows an example in which an evaluation index determined by linear calculation of a basic index, that is, Expression (2) is selected as an objective function. Columns 721 to 723 indicate basic index items, and column 724 indicates threshold values of the items (boundary values of the range calculated by the driving range extracting unit 111). The product of the reciprocal of the column 725 and the column 726 becomes the coefficient w_i for the basic index h_i corresponding to the equation (2).

  Further, the configuration / plan display unit 153 may accept, as an input, correction of parameters such as threshold values and weights in the objective function display panel 711. When the correction is received, the configuration / plan display unit 153 transmits the corrected parameter to the operation plan planning unit 114. On the other hand, the operation planning apparatus 101 presents the extracted (experienced) operator's past requests and judgments to the operator through the processing in the configuration / plan display unit 153 described above, and as necessary. Amendments can be accepted.

  FIG. 8 is a diagram illustrating an example of an operation plan display window 801. With reference to FIG. 8, a configuration in which the configuration / plan display unit 153 presents the operation plan and the evaluation result related to the basic index to the operator on the display unit 206 or the like will be described.

  The operation plan display window 801 displayed by the configuration / plan display unit 153 on the display unit 206 such as a display in this case includes an item selection panel 811, an index value display panel 812, and a plan detail display panel 813.

  The configuration / plan display unit 153 displays the operation plan and basic index item options to be displayed on the item selection panel 811 among the above-described panels. Further, the configuration / plan display unit 153 receives an input of an operation plan and basic index items to be displayed in response to a selection operation from the operator on the item selection panel 811.

  In addition, the configuration / plan display unit 153 positions the operation plan in the distribution of the operation range extracted by the operation range extraction unit 111 for the item selected by the item selection panel 811 with respect to the index value display panel 812. Display and calculate and display the corresponding basic index.

  In addition, the configuration / plan display unit 153 displays the index function 821 and the threshold value 822 received from the operation range extraction unit 111 in a graph on the index value display panel 812, and superimposes the plan value 825 on the index function 821 and the threshold value 822. , And the deviation degree 827 from the operation range is displayed as a numerical value.

  The configuration / plan display unit 153 displays, for the plan detail display panel 813, the transition in time series in the plan period of the item selected by the item selection panel 811 among the operation plans received from the configuration / plan display unit 153. A graph is displayed as the planned value time series 835. Further, the configuration / plan display unit 153 displays the threshold value 831 in an overlapping manner.

The operation plan planning apparatus 101 evaluates the planned operation plan from the viewpoint of satisfaction (deviation degree) of the extracted demand and judgment of the operator by the processing of the configuration / plan display unit 153 described above. Can be presented against.

--- Example 2 ---
In the present embodiment, an example in which the operation planning apparatus 101 includes an operation condition extraction unit 915 that extracts the operation condition of the plant based on the device state of the monitoring control system 102 in addition to the configuration of the first embodiment will be described.

  In this case, the operation plan planning apparatus 101 searches for and calculates a change that the problem configuration learning unit 112 adds to the configuration of the evaluation index for each operation condition, and the operation plan planning unit 114 calculates the evaluation index corresponding to the operation condition. The configuration will be selected as the objective function. Since most of the configuration and processing in the operation planning apparatus 101 are the same as those in the first embodiment described above, the overlapping description will not be repeated and only the differences will be described.

  FIG. 9 is a diagram illustrating a logical configuration of the operation plan planning apparatus 101 according to the second embodiment. Specifically, the configuration example of the operation control system 100 including the operation plan planning apparatus 101 including the operation condition extraction unit 915 is illustrated. FIG.

  The operation condition extraction unit 915 in the present embodiment acquires the operation result data from the operation result data storage unit 121, extracts the difference from the normal state of the equipment state in the water plant as the operation condition from the operation result data, The operation condition is transmitted to the problem configuration learning unit 112 and the operation plan planning unit 114.

  For example, when the operation result data includes information that the power distribution facility and the equipment of the distribution pump facility have temporarily failed due to electrical inspection or the like in a certain period, the operation condition extraction unit 915 And outage in the distribution pump facility is extracted as one operation condition. The operation conditions extracted by the operation condition extraction unit 915 are not limited to power outages, such as failure of mechanical equipment such as pumps, suspension of use, failure of instrumentation equipment such as flowmeters, and some functional limitations of the monitoring control system. Including.

  The problem configuration learning unit 112 according to the present embodiment additionally receives the operation condition from the operation condition extraction unit 915 described above, learns the problem configuration for the normal condition and each operation condition, and converts the problem configuration into the operation plan planning unit 114 and the configuration. -It transmits to the plan display part 153. Details of the problem configuration learning unit 112 of this embodiment will be described later with reference to FIG.

  The operation planning unit 114 according to the present embodiment additionally receives the operation conditions in the planning target period from the operation condition extraction unit 915, and the problem configuration corresponding to the operation conditions, particularly the objective function, from the problem configuration learning unit 112. The operation plan is prepared using the problem configuration corresponding to the operation condition, and the operation plan is transmitted to the plan output unit 152 and the configuration / plan display unit 153.

  FIG. 10 is a diagram illustrating an example of a flow of an operation plan planning method according to the present embodiment, and more specifically, a flowchart illustrating processing of the problem configuration learning unit 112. Hereinafter, the processing of the problem configuration learning unit 112 in this embodiment will be described with reference to FIG.

  First, the problem configuration learning unit 112 starts processing (step 1001). This start is executed by receiving a predetermined instruction from a predetermined device via the input unit 205 or the communication control unit 204, or detecting the arrival of a predetermined time.

Next, the problem composition learning unit 112 receives the operation result data from the operation result data storage unit 121, the list of basic indexes and the calculation method from the operation range extraction unit 111, and the past operation from the operation condition extraction unit 915. Each condition is received (step 1002).

  Next, the problem configuration learning unit 112 learns the problem configuration for the normal condition, that is, the operation condition that is the most typical and has a long operation period (step 1003). In this step, the processing period from step 603 to step 606 in FIG. 6 is performed by limiting the learning target period p to a period that is a normal condition.

  Next, the problem configuration learning unit 112 learns the problem configuration for each operation condition other than the normal condition (step 1004). In this step, the learning target period p is limited to the period corresponding to each operation condition, and the configuration of the evaluation index is limited to the configuration close to the index of the normal condition. Processing up to 606 is performed.

  Note that the above-described configuration of the evaluation index is a configuration close to that of the normal condition, for example, a vector of constants (w_i and k_i in the expressions (1) and (2)) that determine the evaluation index of the operation condition. And a constant vector that defines an evaluation index under normal conditions is smaller than a predetermined value. For the distance between the vectors, the L1 norm of the difference between the two vectors, the L2 norm, the number of different elements, or the like can be used.

  Next, the problem configuration learning unit 112 transmits the normal condition and the problem configuration for each operation condition obtained in steps 1003 and 1004 described above to the operation plan planning unit 114 and the configuration / plan display unit 153 (step 1005). ), The flow ends (step 1006).

  The operation plan planning apparatus 101 according to the present embodiment has a device condition in the water plant having a special operation condition by the additional processing in the operation condition extraction unit 915, the problem configuration learning unit 112, and the operation plan planning unit 114 described above. Thus, it is possible to extract changes in the operator's past requests and judgments regarding operation, and to create an operation plan according to the state of the equipment.

  FIG. 11 is a diagram showing an example of the problem condition display window 1101 by operation condition. With reference to FIG. 11, a configuration in which the configuration / plan display unit 153 presents the problem configuration for each operation condition to the operator will be described.

  The operation condition-specific problem configuration display window 1101 displayed on the display unit 206 such as a display by the configuration / plan display unit 153 includes an operation condition selection box 1110, a determination criterion display panel 1120, and a configuration change display panel 1130.

  The configuration / plan display unit 153 displays a list of the operation conditions other than normal in the operation condition selection box 1110 out of the above so as to be selectable by a pull-down menu or the like, and allows the operator to select the operation condition to be displayed. , Accept as input. In FIG. 11, the “reservoir C facility power outage”, that is, the condition that the facility of the water distribution pump facility installed in the water reservoir and the distribution reservoir C has failed is selected by the operator.

  Further, the configuration / plan display unit 153 displays a determination criterion related to the operation condition selected in the operation condition selection box 1110 described above on the determination criterion display panel 1120. In the table of the determination criterion display panel 1120, a column 1121 displays data items used for determination, and a column 1122 displays conditions for determining that the above-described operation conditions are satisfied with respect to the data items.

Further, the configuration / plan display unit 153 displays a change in the problem configuration to be added to the normal problem configuration in the operation condition selected in the operation condition selection box 1110 described above on the configuration change display panel 1130. . In the example of FIG. 11, columns 1131 to 1133 of the table indicate items of basic indicators, and columns 1134 to 1136 indicate numerical values to be changed.

  The operation plan planning apparatus 101 of the present embodiment presents the operator's past requests and changes in judgments extracted for each operation condition to the operator by the processing of the configuration / plan display unit 153 described above, and is necessary. Corrections can be accepted accordingly.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  Each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files for realizing each function can be stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

  According to this embodiment, it is possible to make an operation plan that appropriately reflects the individual plant situation and operation know-how.

  At least the following will be clarified by the description of the present specification. That is, in the operation planning apparatus of the present embodiment, the problem configuration learning unit may generate the evaluation index by combining the basic index by linear calculation and maximum value calculation.

  According to this, it is possible to efficiently generate an appropriate evaluation index for planning an operation plan.

  Further, in the operation plan planning apparatus of the present embodiment, the problem configuration learning unit may use an index of robustness of the operation plan with respect to an error in demand prediction as a basic index.

  According to this, even when the actual demand is greatly deviated with respect to the demand prediction, it can be considered as a main point of the operation plan that there is no problem in the actual plant operation correspondingly.

  The operation plan planning apparatus according to the present embodiment further includes an operation condition extraction unit that extracts an operation condition of the plant based on a state of a predetermined device in the plant indicated by the operation result data, and the problem configuration learning The unit searches for changes to be made to the configuration of the evaluation index for each operation condition, and the operation plan planning unit selects the configuration of the evaluation index corresponding to the operation condition as an objective function. It is good.

  According to this, it is possible to efficiently search for an evaluation index.

  The operation plan planning apparatus of the present embodiment may further include a configuration / plan display unit that displays the operation plan and a deviation degree of the operation plan from the operation range on a predetermined device.

  According to this, the user of the operation plan planning apparatus can easily recognize the planned operation plan and its contents, and can easily understand the quality and the like.

  The operation plan planning apparatus according to the present embodiment may further include a configuration / plan display unit that displays a configuration of the searched evaluation index on a predetermined device and receives a correction instruction for the configuration.

  According to this, it becomes possible to correct the evaluation index to be in accordance with the user's thoughts or actual situation.

  Further, in the operation planning apparatus of the present embodiment, the water supply / distribution process of the water plant is the target of the operation plan, and the operation range extraction unit is configured to extract the water intake, the water supply amount, the water distribution amount, the distribution reservoir water level, the pump and the valve. The distribution may be extracted with respect to at least one of the number of operation switching operations, the interval, and the time zone.

  According to this, in the operation plan regarding the water plant, it is possible to specify a basic index regarding an event that can be a key, and thereby to improve the evaluation index using this and the accuracy in the operation plan.

  Further, in the operation plan planning apparatus of the present embodiment, the water supply / distribution process of the water plant is the target of the operation plan, and the problem configuration learning unit determines the degree of coincidence between the operation plan draft and the operation result data relating to the period. It is also possible to calculate based on at least one of water intake amount, water supply amount, water distribution amount, water distribution reservoir water level, and pump and valve operation switching time.

  According to this, in the operation plan regarding the water plant, it is possible to specify an evaluation index with good accuracy with respect to an event that can be a key.

DESCRIPTION OF SYMBOLS 100 Operation control system 101 Operation plan planning apparatus 102 Monitoring control apparatus 103 Sensor 104 Sub controller 111 Operation range extraction part 112 Problem structure learning part 113 Demand prediction part 114 Operation plan planning part 121 Operation performance data storage part 151 Data collection part 152 Plan output Unit 153 Configuration / plan display unit 201 CPU
202 Memory 203 Media Input / Output Unit 204 Communication Control Unit 205 Input Unit 206 Display Unit 207 Peripheral Device IF Unit 210 Bus 220 Network

Claims (10)

A device for planning a plant operation plan,
An operation range extraction unit that extracts a distribution of an operation range related to a predetermined event from operation result data related to the plant, and that defines a basic index indicating a deviation degree from the operation range for each predetermined event;
By combining the basic indicators of the predetermined event, generating an evaluation indicator in the operation plan, and solving the optimization problem with the constraint condition regarding the configuration of the plant and the evaluation indicator as an objective function, the past predetermined period Formulating an operation plan, calculating a degree of coincidence between the operation plan and the operation result data relating to the period, and a problem configuration learning unit that searches for a configuration of the evaluation index that maximizes the degree of coincidence,
Based on a predetermined demand forecast value related to the plant, an operation plan drafting unit that formulates an operation plan by solving an optimization problem with the searched evaluation index as an objective function;
An operation planning apparatus characterized by comprising: The problem configuration learning unit
2. The operation planning apparatus according to claim 1, wherein the evaluation index is generated by combining the basic index by linear calculation and maximum value calculation. The problem configuration learning unit
2. The operation plan drafting apparatus according to claim 1, wherein an index of robustness of the operation plan with respect to an error in demand prediction is used as a basic index. Based on the state of the predetermined equipment in the plant indicated by the operation result data, further comprising an operation condition extraction unit that extracts an operation condition of the plant,
The problem configuration learning unit searches for a change to be made to the configuration of the evaluation index for each operation condition,
The operation planning unit selects a configuration of an evaluation index corresponding to the operation condition as an objective function.
The operation planning apparatus according to claim 1, wherein: Further comprising a configuration / plan display unit for displaying the operation plan and a deviation degree from the operation range of the operation plan on a predetermined device,
The operation planning apparatus according to claim 1, wherein:   The operation plan planning apparatus according to claim 1, further comprising a configuration / plan display unit that displays a configuration of the searched evaluation index on a predetermined device and receives a correction instruction for the configuration. The water supply / distribution process of the water plant is the target of the operation plan,
The operation range extraction unit extracts the distribution for at least one of water intake, water supply, water distribution, distribution reservoir water level, number of operation switching operations of pumps and valves, interval, and time zone.
The operation planning apparatus according to claim 1, wherein: The water supply / distribution process of the water plant is the target of the operation plan,
The problem composition learning unit determines the degree of coincidence between the operation plan and the operation result data related to the period, at least one of the intake water amount, the water supply amount, the water distribution amount, the water reservoir level, and the difference between the pump and valve operation switching times. Is calculated based on
The operation planning apparatus according to claim 1, wherein: An operation control system for controlling a plant based on an operation plan,
A monitoring control device that collects measurement values from the sensors of the plant and gives a control instruction to a sub-controller that controls the plant;
An operation plan drafting device for drafting an operation plan as a control target of the monitoring control device,
Operation that extracts a distribution of an operation range related to a predetermined event from operation result data that is a measurement value related to the plant obtained through the monitoring and control device, and sets a basic index indicating a deviation from the operation range for each predetermined event. A range extractor;
By combining the basic indicators of the predetermined event, generating an evaluation indicator in the operation plan, and solving the optimization problem with the constraint condition regarding the configuration of the plant and the evaluation indicator as an objective function, the past predetermined period Formulating an operation plan, calculating a degree of coincidence between the operation plan and the operation result data relating to the period, and a problem configuration learning unit that searches for a configuration of the evaluation index that maximizes the degree of coincidence,
Based on a predetermined demand forecast value related to the plant, an operation plan drafting unit that formulates an operation plan by solving an optimization problem with the searched evaluation index as an objective function;
An operation planning device comprising:
An operation control system characterized by including: An information processing device that creates a plant operation plan
An operation range extraction process for extracting a distribution of an operation range related to a predetermined event from operation result data related to the plant, and defining a basic index for each predetermined event indicating a deviation degree from the operation range;
By combining the basic indicators of the predetermined event, generating an evaluation indicator in the operation plan, and solving the optimization problem with the constraint condition regarding the configuration of the plant and the evaluation indicator as an objective function, the past predetermined period Drafting an operation plan, calculating a degree of coincidence between the operation plan draft and the operation result data relating to the period, and a problem configuration learning process for searching for a configuration of the evaluation index that maximizes the degree of coincidence;
Based on a predetermined demand forecast value related to the plant, an operation plan drafting process for drafting an operation plan by solving an optimization problem with the searched evaluation index as an objective function;
An operation planning method characterized by executing

Images