Water supply scheduling method and equipment based on expert knowledge and annealing evolutionary algorithm, and medium

Abstract

The invention relates to a water supply scheduling method and equipment based on expert knowledge and an annealing evolutionary algorithm, and a medium. The method comprises the following steps: acquiring real-time environment data and water supply related data; searching a water supply scheduling expert knowledge base constructed according to historical data according to the real-time environment data and the water supply related data, and searching a closest historical working condition as an initial scheduling working condition; an initial scheduling working condition is used as a first generation population, a simulated annealing algorithm is adopted to carry out population evolution, local reasoning is carried out through heredity, crossover, variation and annealing of each population by utilizing a dynamic planning thought, scheme optimization is carried out according to a target fitness function, an optimal water supply scheduling scheme is determined, and each individual represents a water supply scheduling scheme. And representing the on-off state matrix of all water pumps or valves at the current time. Compared with the prior art, the method has the advantages that the scheduling scheme solving efficiency is high, and the fitness function does not need to be continuously and manually adjusted.

Description

Technical Field

[0001] The present invention relates to the field of water supply scheduling decision-making, and in particular to a water supply scheduling method, equipment and medium based on expert knowledge and annealing evolutionary algorithm. Background Art

[0002] Water supply scheduling needs to comprehensively consider the water supply capacity of each plant, user pressure, water storage capacity and water inlet and outlet conditions of the pool, user water consumption and water balance between different water supply areas. With the development of the city, the difficulty of water supply scheduling is increasing and it is also facing greater challenges. At present, most of the scheduling is done by manual experience, and different dispatchers have different scheduling methods. New employees need a long time to be competent for the complex scheduling. In order to better improve the scheduling and handling capabilities and standardize the scheduling work, it is increasingly important to formulate a scientific and reasonable scheduling plan.

[0003] At present, many scholars at home and abroad mainly focus on the improvement and development of various optimization scheduling algorithms, that is, to improve the feasibility of theory. Although there are some successful examples of optimization scheduling, the promotion and application of various optimization scheduling methods are still greatly restricted, that is to say, its practicality has certain limitations.

[0004] The microscopic model is mainly used in the pipeline network operating model. However, since it is time-consuming to solve the microscopic model of a large pipeline network, it is necessary to adopt a compromise between the computational efficiency and the number of scheduling period divisions when optimizing the scheduling decision.

[0005] In recent years, various intelligent algorithms have been proposed, such as genetic algorithm (GA), particle swarm algorithm (PSO), greedy algorithm (Greedy Algorithm), ant algorithm (Ant Colony Algorithm), etc., which provide new ideas for solving large-scale nonlinear mixed discrete variable optimization problems. Although there have been studies on this application, it is mainly tentative or only optimizes the calculation of the optimal combination of water pumps in the pumping station without considering the pipeline network, or only calculates the optimal water pump operation combination in one period, and cannot solve the water storage scheduling system that requires continuous period calculation.

[0006] Expert knowledge mainly includes knowledge from books and knowledge gained by scheduling experts in long-term work practice. There are many uncertainties in knowledge, which need to be further screened to form effective expert knowledge. Using expert knowledge for intelligent scheduling has the characteristics of rapid implementation, but it also often faces problems such as poor flexibility, difficulty in optimization and updating, and limited accuracy.

[0007] The genetic simulated annealing algorithm is an optimization algorithm that combines the genetic algorithm and the simulated annealing algorithm. It combines the global fast convergence characteristics of the genetic algorithm in the search and the local search ability of the simulated annealing algorithm to improve the efficiency and accuracy of solving optimization problems. However, when faced with large-scale, multi-period nonlinear mixed discrete variable optimization problems, the genetic simulated annealing algorithm requires huge computing power and time on the one hand, and on the other hand, the formulation of the fitness function requires very experienced water supply and drainage and software engineers to continuously adjust and debug according to the actual situation. It has poor controllability and very high requirements for personnel. Summary of the invention

[0008] The purpose of the present invention is to provide a water supply scheduling method, equipment and medium based on expert knowledge and annealing evolutionary algorithm. The expert knowledge base is used to enable the simulated annealing algorithm to calculate and converge more quickly when processing a large amount of data, thereby improving the computational efficiency. At the same time, the fitness function can be adjusted according to actual constraints, and a reliable and effective water supply scheduling plan can be obtained without manual design.

[0009] The purpose of the present invention can be achieved by the following technical solutions:

[0010] According to a first aspect of the present invention, a water supply scheduling method based on expert knowledge and annealing evolutionary algorithm is provided, comprising the following steps:

[0011] Access real-time environmental and water supply-related data;

[0012] Search the water supply scheduling expert knowledge base built based on historical data based on real-time environmental data and water supply-related data to find the closest historical operating conditions as the initial scheduling conditions;

[0013] The initial scheduling condition is taken as the first generation population, and the simulated annealing evolutionary algorithm is used for population evolution. Through the inheritance, crossover, mutation, and annealing of each population, local reasoning is performed using the dynamic programming idea, and the scheme is optimized according to the target fitness function to determine the optimal water supply scheduling scheme. Each gene represents the quantitative adjustment of a water pump or valve, and each individual represents a water supply scheduling scheme, which is represented by the switch state matrix of all water pumps or valves at the current time.

[0014] The environmental data includes weather data, and the water supply related data includes the liquid level of the clear water tank, the volume of the water supply line, the pump start status of the water plant / pump station, the water output, the water consumption and the water demand of each area.

[0015] The water output is predicted based on a neural network, specifically: utilizing the nonlinear mapping characteristics of the neural network, establishing an implicit nonlinear model of multiple factors including the water pump status of the water plant pump station, the pressure at the pressure measuring point of the pipe network, the liquid level of the downstream water tank and the water output of the pump station, and the pressure at the previous moment of the pressure measuring point, and predicting the water output through the above factors.

[0016] The construction process of the water supply scheduling expert knowledge base is as follows:

[0017] Data preprocessing: Process the daily scheduling data according to the data cleaning rules, remove and supplement the entries with short-term data missing and data anomalies. If the data anomalies in the scheduling data of the day exceed the preset value, the date will be screened out in the sample data; if special working conditions are encountered, manual marking will be performed, and instructions will be added or discarded;

[0018] Feature set induction: The pre-processed scheduling plan data is summarized into feature sets to form a water supply scheduling expert knowledge plan library. The features in the feature set include total water consumption, regional water consumption, weather, clear water tank level, volume of different water supply lines, scheduling times, and instructions.

[0019] The objective fitness function is constructed based on the safety principle and the economic principle, wherein the safety principle is measured by the water tank liquid level, the total water storage volume, the water volume of the water plant pump station, the pipe network pressure, and the water pump operation constraints, and is expressed in the form of a loss function; the economic principle is expressed by the water pump energy consumption, and the water pump energy consumption is calculated by the actual power of the water pump combination scheme.

[0020] The target fitness function is expressed as:

[0021]

[0022] Where: f represents the target fitness function; L n (t) represents the loss function of the nth constraint in the tth time period, N is the total number of constraints; P i (t) represents the actual pump power of pump No. i in time period t; T represents the total number of time periods.

[0023] The constraints include the water supply and demand balance constraint of the total pipeline network, the water supply and demand balance constraint of the pipeline network in the independent water supply area, the total volume constraint of the water tank in different water supply ranges, the water tank volume constraint, the water tank liquid level constraint, the pipeline monitoring point / plant station outlet pressure constraint, the plant station water volume constraint, the pump operation efficiency constraint and the pump operation constraint; the loss function corresponding to the water supply and demand balance constraint of the total pipeline network and the water supply and demand balance constraint of the pipeline network in the independent water supply area is determined according to the degree to which the pipeline flow deviates from the supply and demand balance value, and the loss function corresponding to the total volume constraint of the water tank in different water supply ranges, the water tank volume constraint, the water tank liquid level constraint, the pipeline monitoring point / plant station outlet pressure constraint, the plant station water volume constraint, the pump operation efficiency constraint and the pump operation constraint is determined according to the degree of deviation between the variables in different constraints and their corresponding upper and lower boundary conditions. The greater the deviation between the variable value and the upper and lower boundaries, the greater the loss function.

[0024] The method for determining the boundary conditions is: collecting historical data and scheduling experience data, analyzing the water plant's production capacity, hourly water output, pool liquid level, pump station operation, inter-regional scheduling valves, pressure, and pool volume in different areas, determining the reasonable range of each variable, and using the boundary value of the reasonable range as the boundary condition.

[0025] According to a second aspect of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0026] According to a third aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, wherein the program implements the method described above when executed by a processor.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. The present invention uses an expert knowledge base to search for historically similar optimal operating conditions in the expert knowledge base based on current real-time data, as the initial scheduling conditions, to generate primary genes, and to slide or not operate according to the pump-on time of the primary genes, so that the simulated annealing evolutionary algorithm can calculate and converge more quickly when processing a large amount of data.

[0029] 2. Combine historical data and dispatching experience to fit various boundary problem solution plans, that is, in the case of abnormal water level, abnormal pressure, etc., set local problem solutions based on historical data or manual experience, aiming to guide the algorithm to solve local problems. In the process of population generation evolution, for individuals encountering problems, new individuals are generated by comparing various plans to quickly guide the evolution of the group.

[0030] 3. The present invention proposes for the first time the use of simulated annealing evolution method in scheduling scheme optimization, which combines simulated annealing and genetic algorithm with complementary advantages. It can effectively prevent the problems of poor global search ability and low efficiency of simulated annealing algorithm and poor local search ability and premature phenomenon of genetic algorithm, and ultimately improve the efficiency of solving optimization problems with high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is a flow chart of the method of the present invention;

[0032] Figure 2 It is a schematic diagram of the kernel optimization strategy based on CPU characteristics of the present invention. DETAILED DESCRIPTION

[0033] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0034] This embodiment provides a water supply scheduling method based on expert knowledge and annealing evolutionary algorithm, with the goals of stable operation of the pipeline network, safe and controllable power plants, sufficient overall water pool volume, less pump scheduling times, and optimized energy consumption per thousand tons of water. The most appropriate scheduling scheme is selected to achieve a balance between water supply and demand in the plant network and achieve the goal of scheduling management.

[0035] like Figure 1 As shown, the method comprises the following steps:

[0036] S1, obtain real-time environmental data and water supply related data.

[0037] In this embodiment, the environmental data includes weather data, and the water supply related data includes the liquid level of the clear water tank, the volume of the water supply line, the pumping status of the water plant / pump station, the water output, the water consumption and the water demand of each area.

[0038] Among them, the water output is predicted based on the neural network, specifically: using the nonlinear mapping characteristics of the neural network, an implicit nonlinear model is established including the pump status of the pump station, the pressure of the pipe network pressure measuring point, the liquid level of the downstream pool and the water output of the pump station, the pressure of the pressure measuring point at the previous moment, etc., and the water output is predicted through the above factors. Since the model has no specific functional relationship, its implicit mapping method is obtained by training and learning historical data, which has strong versatility and adaptability.

[0039] For water consumption, we take the historical water consumption and weather of different regions as sample data to train the neural network model. After the training is completed, we input the weather at that time of day to predict the water consumption at different times of the day and in different regions.

[0040] S2, based on real-time environmental data and water supply related data, searches the water supply scheduling expert knowledge base built based on historical data to find the closest historical operating condition as the initial scheduling condition.

[0041] The construction process of the water supply scheduling expert knowledge base is as follows:

[0042] Data preprocessing: Process the daily scheduling data according to the data cleaning rules, remove and supplement the entries with short-term data missing and data anomalies. If the data anomalies in the scheduling data of the day exceed the preset value, delete the date in the sample data; if special working conditions are encountered, manually mark them, add instructions or discard them;

[0043] Feature set induction: The pre-processed scheduling plan data is summarized into feature sets to form a water supply scheduling expert knowledge plan library. The features in the feature set include total water consumption, regional water consumption, weather, clear water tank level, volume of different water supply lines, scheduling times, and instructions.

[0044] In this step, the boundary conditions are also sorted out:

[0045] Collect historical data and track the dispatching experience of each dispatcher, analyze the constraints such as water plant capacity, hourly water output, pool liquid level, pump station operation, inter-regional dispatching valves, pressure, and volume of different areas, and determine the reasonable range as the boundary condition for subsequent dispatching; at the same time, according to the actual situation in the pipeline network, sort out the corresponding rules and restrictions, such as dispatching preferences, periodic maintenance operations, etc.

[0046] S3, taking the initial scheduling condition as the first generation population, adopts simulated annealing evolutionary algorithm for population evolution, through the inheritance, crossover, mutation, annealing of each population, uses dynamic programming ideas for local reasoning, and selects the scheme according to the target fitness function to determine the optimal water supply scheduling scheme, where each gene represents the quantitative regulation of a water pump or valve, and each individual represents a water supply scheduling scheme, which is represented by the switch state matrix of all water pumps or valves at the current time.

[0047] In this embodiment, the target fitness function is constructed based on the safety principle and the economic principle, wherein the safety principle is measured by the water tank liquid level, the total water storage volume, the water volume of the water plant pump station, the pipe network pressure, and the water pump operation constraints, and is expressed in the form of a loss function; the economic principle is expressed by the water pump energy consumption, and the water pump energy consumption is calculated by the actual power of the water pump combination scheme.

[0048] The target fitness function is expressed as:

[0049]

[0050] Where: f represents the target fitness function; L n (t) represents the loss function of the nth constraint in the tth time period, N is the total number of constraints; P i (t) represents the actual pump power of pump No. i in time period t; T represents the total number of time periods.

[0051] Among them, the constraints include the water supply and demand balance constraint of the total pipeline network, the water supply and demand balance constraint of the pipeline network in the independent water supply area, the total volume constraint of the water tank in different water supply ranges, the water tank volume constraint, the water tank liquid level constraint, the pipeline monitoring point / plant station outlet pressure constraint, the plant station water volume constraint, the pump operation efficiency constraint and the pump operation constraint. The loss function corresponding to the water supply and demand balance constraint of the total pipeline network and the water supply and demand balance constraint of the pipeline network in the independent water supply area is determined according to the degree to which the pipeline flow deviates from the supply and demand balance value. The loss function corresponding to the total volume constraint of the water tank in different water supply ranges, the water tank volume constraint, the water tank liquid level constraint, the pipeline monitoring point / plant station outlet pressure constraint, the plant station water volume constraint, the pump operation efficiency constraint and the pump operation constraint is determined according to the degree of deviation between the variables in different constraints and their corresponding upper and lower boundary conditions. The greater the deviation between the variable value and the upper and lower boundaries, the greater the loss function. Specifically, the loss function is expressed as follows:

[0052] 1. Water supply and demand balance of the main pipe network:

[0053] L1(t)=f1(∑Q j -(∑D K +∑ΔV K ))

[0054] 2. Constraints on water supply and demand balance in the pipe network of independent water supply areas

[0055] L2(t)=f2(∑Q j，k -(D K +∑ΔV i,K ))

[0056] 3. Constraints on total volume of water tanks in different water supply ranges

[0057] L3(t)=f3(ΣV Kmin ,ΣV K ,∑V Kmax ),∑V Kmin ≤∑V K ≤ΣV Kmax

[0058] 4. Pool volume constraints

[0059] L4(t)=f4(V jmin ,V j ,V jmax ),V jmin ≤V j ≤Vjmax

[0060] 5. Pool level constraints

[0061] L5(t)=f5(Z jmin ,Z j ,Z jmax ),Z jmin ≤Z j ≤Z jmax

[0062] 6. Pipeline network monitoring point / plant outlet pressure constraints

[0063] L6(t)=f6(H jmin ,H j ,H jmax ),H jmin ≤H j ≤H jmax

[0064] 7. Water volume constraints at power plants and stations

[0065] L7(t)=f7(Q jmin ,Q j ,Q jmax ),Q jmin ≤Q j ≤Q jmax

[0066] 8. Pump operation efficiency constraints

[0067] L8(t)=f8(η imin ,η i ,η imax ),η imin ≤η i ≤η imax

[0068] 9. Pump operation constraints

[0069] L9(t)=f9(n jmin ,n j ,n jmax ),n jmin ≤n j ≤n jmax

[0070] Where: Q is the pipe network flow; P is the pipe network pressure; η i is the operating efficiency of the pump No. i; n i is the speed of the i-th pump; ∑Q j is the total water supply; ∑D K is the total water demand; i, j are the measurement point numbers or plant station numbers or water tank numbers or water supply range numbers; K is the independent water supply area number; Z jmin, Z jmax is the lower upper limit of the liquid level constraint; V jmin , V jmax is the lower upper limit of the volume constraint; H jmin , H jmax is the lower upper limit of the water pressure constraint; Q jmin , Q jmax is the lower upper limit of water quantity constraint; η jmin , η jmax is the upper limit of water pump efficiency; n jmin , n jmax f is the upper limit of the speed of the speed regulating pump; n (x) represents the parameter x of the nth constraint and the loss function L n Linear / nonlinear relationship function between .

[0071] When receiving a water supply scheduling request, obtain the weather, clear water tank level, water plant, pump station pumping status, inlet and outlet flow, inlet and outlet pressure, water consumption, etc. in different regions at the current moment, search for historical similar working conditions from the expert knowledge base based on the current real-time data, and score and compare based on the weather, clear water tank level, tank volume, etc., and select the working condition with the highest score as the initial scheduling condition. In the adaptive genetic simulated annealing algorithm, the selection of the first generation of genes will affect the stability and speed of the algorithm. Combined with the law of water supply network scheduling, the historical operation data of the pump station, manual scheduling experience, and boundary conditions are used as training data to generate an ancestral group. Through the above improvements, the reliability of the calculation results can be effectively improved.

[0072] The specific process of generating the plan is as follows:

[0073] Step 1: Based on the initial real-time state, a population of ancestors with higher scores is generated from the knowledge base as the first generation population x0{k}={}(k=1...n).

[0074] Step 2: Population evolution: By combining the inheritance, crossover, mutation, and annealing of each population, and conducting directed reasoning based on expert knowledge, local reasoning is performed using dynamic programming ideas: the search strategy is dynamically adjusted according to the progress of the search to cope with different solution space characteristics. Figure 2 As shown in the figure, the algorithm adopts the idea of ​​dynamic programming to dynamically divide the state, optimize and solve locally. At the same time, based on the kernel optimization strategy of CPU characteristics, a standby order grabbing mechanism is formulated for each working thread to cope with dynamically generated computing tasks.

[0075] Step 3: Optimize the solution according to the convergence criterion objective fitness function, and find the minimum value as the optimal solution. If the minimum value of the solution's objective function meets the requirements, output the solution as the recommended solution; if it does not meet the requirements, return to the second step to continue the evolution.

[0076] In order to verify the performance of the present invention, this embodiment conducts an automatic calculation experiment of a 24-hour water supply scheduling continuous calculation plan on a super-large water supply enterprise. The system includes 10 independently metered water supply areas, a total of 11 including the head office, multi-level scheduling, and more than 100 water pumps. The annealing evolutionary algorithm based on expert knowledge is analyzed and compared with the genetic algorithm and the model-based adaptive scheduling algorithm method. The experiments are trained and tested according to the experimental regulations of the corresponding data sets. From the comparison in Table 1, it can be seen that the simulated annealing evolutionary algorithm quickly gives a solution when the objective function is met, which fully proves that the expert knowledge and simulated annealing evolutionary algorithm have a good application in solution making. The experimental results have a higher recognition rate than the currently published mainstream algorithms.

[0077] Table 1 Experimental data of the dataset

[0078] algorithm Convergence speed accuracy Objective Function Genetic Algorithms 432000s Poor accuracy, 50%, slow speed -- Adaptive dispatch based on hydraulic model 144000s High accuracy, average speed <100 The present invention 300s High and fast <100

[0079] The electronic device of the present invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0080] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0081] The processing unit performs the various methods and processes described above, such as methods S1 to S3. For example, in some embodiments, methods S1 to S3 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 to S3 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1 to S3 in any other appropriate manner (e.g., by means of firmware).

[0082] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0083] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0084] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0085] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning, or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A water supply scheduling method based on expert knowledge and annealing evolutionary algorithm, characterized in that: The following steps are involved: Access real-time environmental and water supply-related data; Search the water supply scheduling expert knowledge base built based on historical data based on real-time environmental data and water supply-related data to find the closest historical operating conditions as the initial scheduling conditions; The initial scheduling condition is taken as the first generation population, and the simulated annealing evolutionary algorithm is used for population evolution. Through the inheritance, crossover, mutation, and annealing of each population, local reasoning is performed using the dynamic programming idea, and the scheme is optimized according to the target fitness function to determine the optimal water supply scheduling scheme. Each gene represents the quantitative adjustment of a water pump or valve, and each individual represents a water supply scheduling scheme, which is represented by the switch state matrix of all water pumps or valves at the current time.

2. A water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 1, characterized in that: The environmental data includes weather data, and the water supply related data includes the liquid level of the clear water tank, the volume of the water supply line, the pump start status of the water plant / pump station, the water output, the water consumption and the water demand of each area.

3. A water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 2, characterized in that: The water output is predicted based on a neural network, specifically: utilizing the nonlinear mapping characteristics of the neural network, establishing an implicit nonlinear model of multiple factors including the water pump status of the water plant pump station, the pressure at the pressure measuring point of the pipe network, the liquid level of the downstream water tank and the water output of the pump station, and the pressure at the previous moment of the pressure measuring point, and predicting the water output through the above factors.

4. The water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 1 is characterized in that: The construction process of the water supply scheduling expert knowledge base is as follows: Data preprocessing: Process the daily scheduling data according to the data cleaning rules, remove and supplement the entries with short-term data missing and data anomalies. If the data anomalies in the scheduling data of the day exceed the preset value, the date will be screened out in the sample data; if special working conditions are encountered, manual marking will be performed, and instructions will be added or discarded; Feature set induction: The pre-processed scheduling plan data is summarized into feature sets to form a water supply scheduling expert knowledge plan library. The features in the feature set include total water consumption, regional water consumption, weather, clear water tank level, volume of different water supply lines, scheduling times, and instructions.

5. The water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 1 is characterized in that: The objective fitness function is constructed based on the safety principle and the economic principle, wherein the safety principle is measured by the water tank liquid level, the total water storage volume, the water volume of the water plant pump station, the pipe network pressure, and the water pump operation constraints, and is expressed in the form of a loss function; the economic principle is expressed by the water pump energy consumption, and the water pump energy consumption is calculated by the actual power of the water pump combination scheme.

6. A water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 5, characterized in that: The target fitness function is expressed as: Where: f represents the target fitness function; L n (t) represents the loss function of the nth constraint in the tth time period, N is the total number of constraints; P i (t) represents the actual pump power of pump No. i in time period t; T represents the total number of time periods.

7. A water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 6, characterized in that: The constraints include the water supply and demand balance constraint of the total pipeline network, the water supply and demand balance constraint of the pipeline network in the independent water supply area, the total volume constraint of the water tank in different water supply ranges, the water tank volume constraint, the water tank liquid level constraint, the pipeline monitoring point / plant station outlet pressure constraint, the plant station water volume constraint, the pump operation efficiency constraint and the pump operation constraint; the loss function corresponding to the water supply and demand balance constraint of the total pipeline network and the water supply and demand balance constraint of the pipeline network in the independent water supply area is determined according to the degree to which the pipeline flow deviates from the supply and demand balance value, and the loss function corresponding to the total volume constraint of the water tank in different water supply ranges, the water tank volume constraint, the water tank liquid level constraint, the pipeline monitoring point / plant station outlet pressure constraint, the plant station water volume constraint, the pump operation efficiency constraint and the pump operation constraint is determined according to the degree of deviation between the variables in different constraints and their corresponding upper and lower boundary conditions. The greater the deviation between the variable value and the upper and lower boundaries, the greater the loss function.

8. A water supply scheduling method based on expert knowledge and annealing evolutionary algorithm according to claim 7, characterized in that: The method for determining the boundary conditions is: collecting historical data and scheduling experience data, analyzing the water plant's production capacity, hourly water output, pool liquid level, pump station operation, inter-regional scheduling valves, pressure, and pool volume in different areas, determining the reasonable range of each variable, and using the boundary value of the reasonable range as the boundary condition.

9. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images