Control method and control system of multi-agent water plant process system

Abstract

The invention relates to a control method and a control system for a multi-agent water plant process system. The method comprises the following steps: constructing a control objective function for maximizing the global revenue of the water plant process system; performing process simulation on the single agent of the water plant process system according to the constraint condition to obtain an optimal control scheme of the single agent; grouping and integrating the optimal control schemes, and performing multi-target iterative optimization on at least one formed multi-agent cooperative control scheme set according to a control target function to obtain a multi-agent cooperative control strategy meeting the control target function; and dynamically adjusting the multi-agent cooperative control strategy according to the real-time change information of the water plant process system, and deploying the adjusted optimal multi-agent cooperative control strategy to the water plant process system. According to the method, the competition and cooperation relationship between the single agents is managed through an innovative coordination mechanism, and it is ensured that the overall harmony and the water production efficiency of the water plant are not damaged while all the single agents pursue the optimization target of the single agents.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to a control method and a control system for a multi-intelligent water plant process system. Background Art

[0002] With the continuous advancement of science and technology, the production control system of the water plant has gradually integrated technologies such as big data, machine learning and image recognition on the basis of traditional automatic control, and developed a more intelligent control system to enable the water production process of the water plant to achieve the effect of energy saving and consumption reduction.

[0003] At present, according to the water production process of the water plant, the existing control systems that can adopt intelligent control technologies generally include: intelligent alum addition system, intelligent chlorination system, intelligent backwashing system, intelligent pump distribution system, intelligent sludge discharge system and water balance of the whole plant. The above-mentioned single intelligent agents generally adopt edge deployment, self-learning and independent operation to achieve certain control goals by controlling their own single intelligent agents. For example, the intelligent alum addition system can reduce the dosage of coagulant while ensuring that the turbidity of the effluent from the sedimentation tank meets the internal control standard, which can generally save 5% to 20% of the drug consumption; the intelligent backwashing system can maintain the healthy operation of the filter tank (or membrane filter tank), reduce production accidents, save 5% to 10% of energy consumption, and reduce a certain amount of recycled water. However, there is a competitive relationship between the single intelligent agents that influence each other. For example, while intelligent alum addition reduces drug consumption, it may lead to an increase in the filter load of the filtration system, thereby reducing the flushing effect of the intelligent backwashing system; and maintaining a high liquid level in the clear water tank will increase the residence time of chlorine in the water, and the residual chlorine in the water will decay with time. The intelligent chlorination system needs to consume more chlorine to maintain the residual chlorine concentration in the water.

[0004] Therefore, technical personnel in this field urgently need a technical solution that can effectively resolve the competition and game relationship between single-agent controls when collaboratively controlling a multi-agent water plant process system. Summary of the invention

[0005] 1. Technical issues to be resolved In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a control method and control system for a multi-agent water plant process system, which solves the technical problem that the existing water plant control system only considers the energy-saving and consumption-reducing effect of a single agent, thereby increasing the operating costs of other single agents.

[0006] (II) Technical solution In order to achieve the above object, the main technical solutions adopted by the present invention include: In a first aspect, an embodiment of the present invention provides a control method for a multi-agent water plant process system, comprising: Taking the minimization of production cost and optimization of water quality of the water plant process system as the dual control objectives, a control objective function for maximizing the global benefits of the water plant process system is constructed; According to the preset constraints, each single agent of the water plant process system is independently simulated to obtain the optimal control solution for each single agent; The optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set, and the multi-agent collaborative control scheme set is subjected to multi-objective iterative optimization including exchange optimality, production optimality and hybrid optimality according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function; According to the real-time change information of the water plant process system, the multi-agent collaborative control strategy is dynamically adjusted, and the adjusted optimal multi-agent collaborative control strategy is deployed to the water plant process system.

[0007] Optionally, with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system, the control objective function for maximizing the global benefit of the water plant process system is constructed, including: Obtain the production cost information and water quality information of the water plant process system. The production cost information includes the water production process cost and the water supply process cost; According to the water supply quality information and the preset water supply quality target, the deviation between the water supply quality and the water supply quality target is obtained; Based on the preset collaborative reward function and interference penalty function between single agents, combined with the production cost information and the deviation between the actual water quality and the water quality target, a control objective function for maximizing the global benefit of the water plant process system is constructed with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system. Among them, the control objective function is: ; In the formula, k 1 is the weight of water quality, k 2 is the weight of water production cost, k 3 is the weight of water supply cost, k 4 is the weight of the collaborative reward function, k 5 is the weight of the interference penalty function, is the deviation between the actual water quality and the water quality target. V punit The cost of water production process, FT in is the water inlet flow rate, V sunit The cost of water supply process, PT is the actual water supply pressure, FT out is the water supply flow rate,R 1 is the collaborative reward function between single agents, R 2 is the interference penalty function between single agents.

[0008] Optionally, an independent process simulation is performed on each single agent of the water plant process system according to preset constraints to obtain the optimal control solution for each single agent, including: According to the process type of each single agent in the water plant process system, the dual control objectives are disassembled to obtain the dual control sub-objectives of minimizing the production cost of each single agent and optimizing the output result; Traverse all dual control sub-goals, and obtain the constraint conditions for each single agent to perform process simulation based on the influencing factors of the single agent operation; Conduct independent process simulation for each single agent of the water plant process system according to the constraints, analyze the simulation output results and process costs, and obtain the optimal control solution for each single agent; The constraint condition expression is: ; In the formula, For a single agent i The minimum dosage of the agent, For a single agent i The dosage of the agent, For a single agent i The maximum dosage of the agent, For a single agent i The minimum liquid level, Represents a single agent i The liquid level, Represents a single agent i The maximum liquid level, For a single agent i Medium Equipment j The number of runs, For a single agent i Medium Equipment j The maximum number of runs, For a single agent i Medium Equipment a and equipment b Only one can be run at a time.

[0009] Optionally, the optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set, and the multi-agent collaborative control scheme set is subjected to multi-objective iterative optimization including exchange optimality, production optimality and hybrid optimality according to the control objective function, and the multi-agent collaborative control strategy that satisfies the control objective function is obtained, including: According to the acquired single-agent process type, the optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set; According to the preset reward and punishment function between single agents, the multi-agent collaborative control solution set is optimized by multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization to obtain the multi-agent collaborative control strategy solution set; According to the control objective function of maximizing the global benefit of the water plant process system, the multi-agent collaborative control strategy solution set is converged to obtain a multi-agent collaborative control strategy that meets the control objective function; Among them, the reward and punishment functions include: collaboration reward function and interference penalty function; The collaborative reward function is: ; In the formula, λ is the weight of the efficient collaboration reward, δ the weight of the reward for inefficient collaboration; The interference penalty function is: R 2=- μ ; In the formula, μ The intensity of the interference penalty.

[0010] Optionally, according to the preset reward and punishment function between single agents, the multi-agent collaborative control solution set is subjected to multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization, and the multi-agent collaborative control strategy solution set is obtained, including: According to the preset reward and punishment function between single agents, the reward and punishment weights between single agents are obtained; Perform multi-objective coordinated optimization on the multi-agent collaborative control solution set, including exchange optimization, production optimization and hybrid optimization, to obtain the initial multi-agent collaborative control strategy solution set; Through the genetic multi-objective optimization algorithm and combined with the reward and punishment weights, the initial multi-agent collaborative control strategy solution set is iteratively updated to obtain the multi-agent collaborative control strategy solution set.

[0011] Optionally, according to the control objective function of maximizing the global benefit of the water plant process system, the multi-agent collaborative control strategy solution set is converged to obtain a multi-agent collaborative control strategy that satisfies the control objective function, including: Traverse each multi-agent collaborative control strategy in the multi-agent collaborative control strategy solution set, and select and obtain a multi-agent collaborative control strategy that meets the set minimum water quality target; According to the control objective function of maximizing the global benefit of the water plant process system, the global benefit data of the water plant process system of each multi-agent collaborative control strategy after the minority selection is obtained; The multi-agent collaborative control strategy with the largest global benefit data of the water plant process system is determined as the multi-agent collaborative control strategy that meets the control objective function.

[0012] Optionally, dynamically adjusting the multi-agent collaborative control strategy according to the acquired real-time change information of the water plant process system, and deploying the adjusted optimal multi-agent collaborative control strategy to the water plant process system includes: Obtain real-time change information of the water plant process system, including water plant environment change information, water supply target change information and production cost change information; According to the real-time change information, the multi-agent collaborative control strategy is dynamically adjusted to obtain the optimal multi-agent collaborative control strategy for the water plant process system; The optimal multi-agent collaborative control strategy is deployed to the water plant process system to make the operating conditions of the water plant process system optimal.

[0013] In a second aspect, an embodiment of the present invention provides a control system for a multi-agent water plant process system, including: The control objective function construction module is used to construct the control objective function of maximizing the global benefit of the water plant process system with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system; The single-agent control optimization module is used to perform independent process simulation on each single-agent of the water plant process system according to preset constraints to obtain the optimal control solution for each single-agent; The multi-agent control optimization module is used to group and integrate the optimal control schemes to form at least one multi-agent collaborative control scheme set, and perform multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization on the multi-agent collaborative control scheme set according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function; The control strategy dynamic adjustment module is used to dynamically adjust the multi-agent collaborative control strategy based on the real-time change information of the water plant process system, and deploy the adjusted optimal multi-agent collaborative control strategy to the water plant process system.

[0014] In a third aspect, an embodiment of the present invention provides a multi-agent water plant process system, including: A water plant process brain module, used to execute the control method steps of a multi-agent water plant process system described above; The single intelligent body module of the water plant is connected to the process brain module of the water plant, and is used to execute the water production process of the water plant when receiving the control instructions of the process brain module of the water plant.

[0015] Optionally, the single intelligent module of the water plant includes: an intelligent alum addition module, an intelligent chlorination module, an intelligent backwashing module, an intelligent sludge discharge module, an intelligent pump distribution module, an intelligent water balance module, an intelligent monitoring module and an equipment management module.

[0016] (III) Beneficial effects The beneficial effects of the present invention are as follows: the control method of a multi-agent water plant process system proposed by the present invention adopts the optimization control goal of maximizing the global benefit of the water plant process system, comprehensively considers the control constraints of each single agent and the advantages of multi-agent linkage, and provides a new set of dynamic optimal collaborative control solutions for the water plant process system through multi-objective coordinated optimization. It manages the competition and cooperation relationship between single agents through an efficient coordination mechanism, ensuring that each single agent will not damage the overall harmony and water production efficiency of the water plant process system while pursuing its own optimization goal.

[0017] In addition, the present invention can also adaptively adjust the collaborative control strategy of the water plant process system according to the operating data of the water plant process system and the external environment change information, and realize more precise and dynamic control by continuously optimizing the control strategy process. It not only significantly improves the stability and intelligence level of the water plant process system operation, but also effectively reduces the energy consumption of the water plant process system operation and improves the resource utilization efficiency, so as to ensure that the indicators of water quality and water quantity can be continuously optimized. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic flow chart of a control method for a multi-agent water plant process system provided by an embodiment of the present invention; Figure 2 A schematic diagram of a process flow of a multi-agent water plant process system provided by an embodiment of the present invention; Figure 3 A schematic diagram of a process flow of a single-agent process simulation provided by an embodiment of the present invention; Figure 4 An optimal solution set diagram of an intelligent alum adding system provided by one embodiment of the present invention; Figure 5 A schematic diagram of a process flow of genetic multi-objective optimization provided by an embodiment of the present invention; Figure 6 An optimal solution set diagram of the optimal solution frontier PF between the intelligent alum adding system and the intelligent backwashing system provided by one embodiment of the present invention; Figure 7 A schematic diagram of a process for converging a solution set of a multi-agent collaborative control strategy provided by an embodiment of the present invention; Figure 8 A schematic diagram of a process for implementing dynamic adjustment of a multi-agent collaborative control strategy provided by an embodiment of the present invention; Fig. 9 A schematic diagram of the composition of a multi-agent water plant process system provided by an embodiment of the present invention; Fig.10 A schematic diagram of a process flow for collaborative control of a multi-agent water plant process system provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0020] Prior to this, in order to facilitate understanding of the technical solution provided by the present invention, some basic information related to the technical solution of the present application is first introduced below, and the solution of the present application can be subsequently adjusted and controlled based on the basic information.

[0021] Multi-agent: A multi-agent system generally refers to a collection of multiple independent agents, each of which operates independently and communicates and coordinates with each other.

[0022] Alum addition process: During the water production process of the water plant, the alum addition process is to add coagulants into the raw water (the coagulant commonly used in water plants is polyaluminium chloride, Poly Aluminium Chloride, referred to as PAC). Under the action of the coagulant, the colloids and fine suspended matter in the water are condensed into flocs and then continuously precipitated, reducing the sensory indicators of water quality such as turbidity and color of the raw water.

[0023] Chlorination process: During the water production process of a water plant, the chlorination process is to add disinfectants into the water (the commonly used disinfectant in water plants is sodium hypochlorite). Chlorination disinfection kills bacteria and deadly microorganisms in the water, and retains a certain residual chlorine content in the water leaving the factory to prevent bacteria from growing in the process of tap water leaving the water plant to the end user.

[0024] Filtration process: During the water production process of the water plant, the filtration process is to remove suspended matter, microorganisms, etc. in the water through sand filters (using quartz stone as filter material), carbon filters (using activated carbon as filter material), membrane filters (using fiber membranes as filter materials) and other process bodies to achieve the purpose of reducing water turbidity.

[0025] Backwashing process: When the filter is filtering, impurities such as suspended matter in the water are retained in the filter material. After a period of time, the filtering effect of the filter will decrease. Backwashing is performed through air flushing, air-water mixed flushing, water flushing, etc. to remove impurities in the filter layer and restore the filtering effect of the filter.

[0026] Intelligent pump matching technology: During the water production process of the water plant, the intelligent pump matching technology combines the pump units in the water supply pump room through different demand strategies (such as water supply pressure demand, water supply flow demand, pump combination efficiency, pump health, etc.) to achieve the best operating effect.

[0027] refer to Figure 1 As shown, a control method for a multi-agent water plant process system proposed in an embodiment of the present invention, the executor of the method of this embodiment can be a water plant process brain module of the multi-agent water plant process system, the water plant process brain module connects each single agent in the multi-agent water plant process system, and the control method includes: taking the minimization of production cost and optimization of water quality of the water plant process system as dual control objectives, constructing a control objective function for maximizing the global benefit of the water plant process system; performing independent process simulation on each single agent of the water plant process system according to preset constraints to obtain the optimal control scheme for each single agent; grouping and integrating the optimal control schemes to form at least one multi-agent collaborative control scheme set, and performing multi-objective iterative optimization of the multi-agent collaborative control scheme set including exchange optimality, production optimality and hybrid optimality according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function; dynamically adjusting the multi-agent collaborative control strategy according to the acquired real-time change information of the water plant process system, and deploying the adjusted optimal multi-agent collaborative control strategy to the water plant process system.

[0028] This embodiment adopts the optimization control goal of maximizing the global benefits of the water plant process system, comprehensively considers the control constraints of each single agent and the advantages of multi-agent linkage, and provides a new set of dynamic optimal collaborative control solutions for the water plant process system through multi-objective coordinated optimization. It manages the competition and cooperation relationship between single agents through an efficient coordination mechanism to ensure that each single agent will not damage the overall harmony and water production efficiency of the water plant process system while pursuing its own optimization goals.

[0029] In addition, this embodiment can also adaptively adjust the collaborative control strategy of the water plant process system based on the operating data of the water plant process system and external environmental change information, and achieve more precise and dynamic control by continuously optimizing the control strategy process. It not only significantly improves the stability and intelligence level of the water plant process system operation, but also effectively reduces the energy consumption of the water plant process system operation and improves resource utilization efficiency, so as to ensure that the indicators of water quality and water quantity can be continuously optimized.

[0030] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0031] Specifically, refer to Figure 1 As shown, a control method for a multi-agent water plant process system proposed in this embodiment includes: S100. Taking the minimization of production cost and optimization of water quality of the water plant process system as the dual control objectives, a control objective function for maximizing the global benefit of the water plant process system is constructed.

[0032] The control objectives of maximizing the global benefits of the water plant process system are to minimize the production costs of the water plant process system and optimize the water quality, where a minimum water quality target is set. Figure 2 As shown in the figure, the production link of the water plant process system adopts a fixed operation mode, or the intelligent control module is not introduced, and its production cost remains basically unchanged. Therefore, the global profit maximization is provided by the sub-agents that constitute the multi-agent.

[0033] In this embodiment, step S100 may include steps S110-S130: S110. Obtaining production cost information and water quality information of the water plant process system, wherein the production cost information includes water production process cost and water supply process cost.

[0034] The production cost includes the cost of the water production process section and the water supply process section. The water supply process section is the water supply pump room, and the remaining process sections are all water production process sections. The cost of each process section is mainly composed of equipment power costs, chemical consumption costs, and equipment usage costs.

[0035] Water production process cost V punit It is the total cost of equipment power cost, reagent consumption cost and equipment use cost (equipment use cost includes equipment maintenance cost and residual value consumption cost) in each water production process section, and its expression is: (1) In formula (1), n is the number of water production process stages, vp i For water production process i Equipment power costs, vm i For water production process i The cost of drug consumption, vd i For water production process i equipment usage fees.

[0036] Water delivery process cost V punit The power cost of the equipment in the water delivery process V s and equipment usage fees V q The total cost is expressed as: V punit = V s + V q (2) S120. Obtain the degree of deviation between the water supply quality and the water supply quality target based on the water supply quality information and the preset water supply quality target.

[0037] Water supply quality information includes water supply pressure, water supply turbidity, water supply residual chlorine and water supply pH. W out The actual water supply pressure PT , Actual water supply turbidity TU 、Actual water supply residual chlorine CL and the actual water supply pH PH composition.

[0038] W out ={ PT , TU , CL , PH} (3) Water supply quality target By water supply pressure target PT tg , water supply turbidity target TU tg , water supply residual chlorine target CL tg and water supply pH targets PH tg composition, (4) Actual water quality W out Water supply quality goals The deviation degree can describe the quality of water supply. The formula for obtaining is: (5) S130. Based on the preset collaborative reward function and interference penalty function between single intelligent agents, combined with the production cost information and the deviation between the actual water supply quality and the water supply quality target, a control objective function for maximizing the global benefit of the water plant process system is constructed with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system.

[0039] Among them, the control objective function is: (6) In formula (6), k 1 is the weight of water quality, k 2 is the weight of water production cost, k 3 is the weight of water supply cost, k 4 is the weight of the collaborative reward function, k 5 is the weight of the interference penalty function, FT in is the water inlet flow rate, FT out is the water supply flow rate, R 1 is the collaborative reward function between single agents, R 2 is the interference penalty function between single agents.

[0040] S200: Perform independent process simulation on each single intelligent agent in the water plant process system according to preset constraints to obtain the optimal control solution for each single intelligent agent.

[0041] Each single agent in the multi-agent water plant process system can achieve the optimal state of its own control effect, especially the self-intelligent regulation under the influence of changes in the relevant influencing factors of the single agent itself. With the help of the ability of the single agent itself, the simulation input of the process is adjusted to simulate the actual process performance, and the optimal solution of the single agent control is solved. It mainly includes simulating the input of different influencing factors, giving the specific output state performance expectations and process cost statistics through the agent itself, and then analyzing the output state results and statistical results to obtain the optimal control plan of the agent. The specific process is as follows Figure 3 .

[0042] In this embodiment, step S200 may include steps S210-S230: S210. Based on the acquired process type of each single intelligent agent in the water plant process system, the dual control objectives are disassembled to obtain the dual control sub-objectives of minimizing the production cost of each single intelligent agent and optimizing the output result.

[0043] Since the process type of each single intelligent body in the water plant process system is different, the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system can be decomposed into different sub-objectives. For example, in the intelligent alum addition system, its dual control sub-objectives are minimum chemical dosage and optimal effluent turbidity; in the intelligent chlorination system, its dual control sub-objectives are minimum chemical dosage and optimal effluent residual chlorine content; in the intelligent backwashing system, its dual control sub-objectives are optimal effluent turbidity and minimum flushing cost; in the intelligent sludge discharge system, its dual control sub-objectives are optimal effluent turbidity and optimal sludge discharge cost; in the intelligent pump distribution system, its dual control sub-objectives are optimal equipment operating cost and optimal pump distribution efficiency.

[0044] S220 traverses all dual control sub-goals, and obtains the constraint conditions for each single agent to perform process simulation based on the influencing factors of the single agent operation.

[0045] Due to the different process types, the influencing factors affecting the operation of single agents are also different. For example, the main influencing factors of the intelligent alum addition system include: water flow, turbidity, temperature, pH, and drug consumption. The main influencing factors of the intelligent chlorination system include: water flow, turbidity, temperature, residual chlorine target, and drug consumption. The main influencing factors of the intelligent backwash system include: water flow, turbidity, backwash cycle, number of backwashes, backwash intensity, and backwash duration. The main influencing factors of the intelligent sludge discharge system include: water flow, turbidity, sludge discharge valve, sludge discharge vehicle, sludge discharge duration, and sludge discharge cycle. The main influencing factors of the intelligent pump distribution system include: network time-sharing water volume, pressure, clear water tank level, pump group matching, and operating frequency. By analyzing the dual control sub-goals of each single agent and the influencing factors of the single agent operation, the constraints for each single agent to perform independent process simulation can be obtained. The expression of the constraints is: (7) In formula (7), For a single agent i The minimum dosage of the agent, For a single agent i The dosage of the agent, For a single agent i The maximum dosage of the agent, For a single agent i The minimum liquid level, Represents a single agent i The liquid level, Represents a single agent i The maximum liquid level, For a single agent i Medium Equipment j The number of runs, For a single agent i Medium Equipment j The maximum number of runs, For a single agent i Medium Equipment a and equipment b Only one can be run at a time.

[0046] S230. Perform independent process simulation on each single intelligent agent in the water plant process system according to the constraint conditions, and analyze the simulation output results and process costs to obtain the optimal control solution for each single intelligent agent.

[0047] In a specific embodiment, according to the constraints of the intelligent alum adding system, several control inputs of the alum adding process simulation are set, and the process simulation results and process costs of each group of control inputs are counted; then, by analyzing the process simulation results and process costs, the following are obtained: Figure 4 The optimal solution set diagram of the intelligent alum adding system is shown in the figure. The solutions (○) on the curve are all the optimal solutions of the intelligent alum adding system following the change of the influent turbidity. The solutions (*) above the curve are the non-optimal solutions of excessive alum addition, and the solutions (×) below the curve are the non-optimal solutions of insufficient alum addition. Finally, according to Figure 4 The optimal solution set curve is used to reverse the control input of the alum adding process simulation to obtain the optimal control solution of the intelligent alum adding system.

[0048] S300. Group and integrate the optimal control schemes to form at least one multi-agent collaborative control scheme set, and perform multi-objective iterative optimization of the multi-agent collaborative control scheme set including exchange optimality, production optimality and hybrid optimality according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function.

[0049] In this embodiment, step S300 may include steps S310-S330: S310. According to the acquired process type of the single agent, the optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set.

[0050] According to the process type of the single agent, the optimal control schemes of the single agent are grouped and integrated to form at least one multi-agent collaborative control scheme set. For example, the optimal control schemes of the intelligent alum addition system and the intelligent backwashing system are integrated to form the first multi-agent collaborative control scheme set for the flocculation sedimentation process; the optimal control schemes of the intelligent alum addition system, the intelligent backwashing system and the intelligent mud discharge system are integrated to form the second multi-agent collaborative control scheme set for the flocculation sedimentation process; the optimal control schemes of the intelligent chlorination system and the intelligent pump distribution system are integrated to form the multi-agent collaborative control scheme set for the chlorination process.

[0051] S320, based on the preset reward and punishment function between single agents, perform multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization on the multi-agent collaborative control solution set to obtain the multi-agent collaborative control strategy solution set. The reward and punishment function includes: collaborative reward function and interference penalty function.

[0052] When multiple agents jointly optimize the operation of the overall water plant, additional collaboration rewards can be given. The collaboration reward function is: (8) In formula (8), λ is the weight of the efficient collaboration reward, δ The weight of the reward for inefficient collaboration.

[0053] Efficient collaboration mainly means that the multi-agent water plant process system can make reasonable decisions based on the overall goal to avoid redundant work and energy waste. The weight of the efficient collaboration reward is between 1-10. For some small-scale optimization or coordination of non-core tasks, the efficient collaboration reward weight is a low weight value (for example, 1-3). For multi-agent collaboration to optimize energy efficiency and reduce costs while meeting the effluent water quality, the efficient collaboration reward weight can be set to a high-range median value (for example, 7-9).

[0054] The inefficient collaboration reward weight is when the agents fail to cooperate effectively, resulting in energy waste and inefficient operation. A certain amount of reward should be deducted to promote the system to adjust the strategy. For some inefficient collaboration scenarios, the inefficient collaboration reward weight is between -1 and -5. If the system operates inefficiently and the water plant equipment operates at high load for a long time, the inefficient collaboration reward weight can be -2 to -4; if the collaborative operation between multiple agents wastes energy, the inefficient collaboration reward weight can be -3 to -5.

[0055] The interference penalty function is: R 2=- μ (9) In formula (9), μ The intensity of the interference penalty.

[0056] Further, step S320 may include steps S321-S323: S321. Obtain reward and punishment weights between single agents according to a preset reward and punishment function between single agents.

[0057] S322. Perform multi-objective coordinated optimization on the multi-agent collaborative control solution set, including exchange optimization, production optimization and hybrid optimization, to obtain the initial multi-agent collaborative control strategy solution set.

[0058] When the exchange is optimal, the multi-agent collaborative control strategy under any combination of the optimal control solutions of the single agent cannot exceed the maximum control benefit effect.

[0059] When production is optimal, the control plan of each single agent is on the production possibility frontier of the configuration.

[0060] When the hybrid optimality is reached, each collaborative control strategy in the solution set of the intelligent collaborative control strategy can reflect the control objective of the water plant process system.

[0061] S323. Through the genetic multi-objective optimization algorithm, combined with the reward and punishment weights, the initial multi-agent collaborative control strategy solution set is iteratively updated to obtain the multi-agent collaborative control strategy solution set.

[0062] In a specific embodiment, reference Figure 5 As shown, the genetic multi-objective optimization algorithm is used to iteratively update the first multi-agent collaborative control solution set of the flocculation sedimentation process integrated with the optimal control solution of the intelligent alum addition system and the intelligent backwashing system. First, the multi-agent collaborative control strategy solution set that maximizes the global benefit of the multi-agent when the water quality meets the standard is calculated through the reward and punishment function to resolve the competitive relationship between the single agents and prevent the single agents from developing in a selfish direction. Then, the genetic multi-objective optimization algorithm is used to iteratively update the initial multi-agent collaborative control strategy solution set including crossover, mutation and merging to generate the offspring solution set of the first multi-agent collaborative control solution set. Finally, when the number of iterations G meets the set termination evolution iteration number Gen, it is determined that the offspring solution set obtained in the last iteration is the collaborative control strategy solution set of the intelligent alum addition system and the intelligent backwashing system. At the same time, the following can also be obtained: Figure 6 The optimal solution set of the optimal solution frontier PF between the intelligent alum adding system and the intelligent backwashing system is shown.

[0063] S330. Based on the control objective function of maximizing the global benefit of the water plant process system, the multi-agent collaborative control strategy solution set is converged to obtain a multi-agent collaborative control strategy that meets the control objective function.

[0064] The solution set of the multi-agent collaborative control strategy needs to be further combined with the control objective function of maximizing the global benefit of the water plant process system to converge the global optimal solution. The specific convergence process is as follows: Figure 7 Further, step S330 may include steps S331-S333: S331. Traverse each multi-agent collaborative control strategy in the multi-agent collaborative control strategy solution set, and select and obtain the multi-agent collaborative control strategy that meets the set minimum water quality target.

[0065] S332. Based on the control objective function of maximizing the global benefit of the water plant process system, the global benefit data of the water plant process system of each multi-agent collaborative control strategy after minority selection is obtained.

[0066] S333. Determine the multi-agent collaborative control strategy that maximizes the global benefit data of the water plant process system as the multi-agent collaborative control strategy that meets the control objective function.

[0067] S400. Dynamically adjust the multi-agent collaborative control strategy based on the acquired real-time change information of the water plant process system, and deploy the adjusted optimal multi-agent collaborative control strategy to the water plant process system.

[0068] When the system influencing factors such as user demand, water environment, water target, cost, weight, etc. change, it is necessary to re-iterate and re-optimize the solution to achieve the dynamic response of the water plant process control system to real-time changes, so as to meet the optimal working conditions of the water plant process system. The dynamic adjustment of the multi-agent collaborative control strategy is implemented as follows: Figure 8 As shown, Figure 8 Where th1, th2…thn are the threshold targets for re-triggers of adjustment of multi-agent collaborative control strategies in the corresponding dimensions.

[0069] In this embodiment, step S400 may include steps S410-S430: S410, obtaining real-time change information of the water plant process system, the real-time change information including water plant environment change information, water supply target change information and production cost change information.

[0070] S420. Dynamically adjust the multi-agent collaborative control strategy based on real-time change information to obtain the optimal multi-agent collaborative control strategy for the water plant process system.

[0071] S430. Deploy the optimal multi-agent collaborative control strategy to the water plant process system to make the operating condition of the water plant process system optimal.

[0072] In addition, this embodiment also provides a control system for a multi-agent water plant process system, including: The control objective function construction module is used to construct a control objective function for maximizing the global benefit of the water plant process system with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system.

[0073] The single-agent control optimization module is used to perform independent process simulation on each single-agent in the water plant process system according to preset constraints to obtain the optimal control solution for each single-agent.

[0074] The multi-agent control optimization module is used to group and integrate the optimal control schemes to form at least one multi-agent collaborative control scheme set, and perform multi-objective iterative optimization of the multi-agent collaborative control scheme set including exchange optimality, production optimality and hybrid optimality according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function.

[0075] The control strategy dynamic adjustment module is used to dynamically adjust the multi-agent collaborative control strategy based on the real-time change information of the water plant process system, and deploy the adjusted optimal multi-agent collaborative control strategy to the water plant process system.

[0076] In addition, this embodiment also provides a multi-agent water plant process system, including: The water plant process brain module is used to execute the control method steps of a multi-agent water plant process system described above.

[0077] The single intelligent agent module of the water plant is connected to the process brain module of the water plant, and is used to execute the water production process of the water plant when receiving the control instructions of the process brain module of the water plant.

[0078] In this embodiment, reference Fig. 9 As shown, the single intelligent module of the water plant includes: intelligent alum addition module, intelligent chlorination module, intelligent backwash module, intelligent sludge discharge module, intelligent pump distribution module, intelligent water balance module, intelligent monitoring module and equipment management module.

[0079] The intelligent alum addition module is deployed in the coagulant addition process section of the water plant. The module can realize fully automatic and precise control of the coagulant dosage according to the specific requirements of the inlet flow, turbidity, temperature, pH and turbidity of the sedimentation tank effluent, thereby ensuring the continuous stability of the sedimentation tank effluent quality and effectively reducing the consumption of chemicals.

[0080] The intelligent chlorination module is deployed in the chlorination process section of the water plant. The module can automatically and accurately control the amount of disinfectant added according to the requirements of water inlet flow, temperature, pH, and water quality, ensuring stable water quality, reducing drug consumption, and reducing disinfection by-products.

[0081] The intelligent backwash module is deployed in the backwash process section of the water plant. The module can automatically and accurately control the backwash cycle and intensity according to parameters such as inlet flow, turbidity, filter water supply turbidity, liquid level, water supply valve opening and head loss, so as to maintain the normal operation of the filter, improve backwash efficiency and reduce energy consumption.

[0082] The intelligent sludge discharge module is deployed in the sludge discharge process section of the flocculation sedimentation tank of the water plant. The module can adjust the sludge discharge mode, sludge discharge cycle and sludge discharge duration in the flocculation zone and sedimentation zone according to parameters such as water inlet flow and turbidity, maintain the healthy and normal operation of the flocculation sedimentation system, improve sludge discharge efficiency, and save energy and reduce consumption.

[0083] The intelligent pump distribution module is deployed in the water supply pump room or lift pump room process section of the water plant. Under the conditions of meeting the water supply pressure, water supply volume and health status of the pump group, the module regulates the pump-machine combination to achieve the optimal operating effect of the pump-machine combination and thus reduce energy consumption.

[0084] Intelligent water balance module, which monitors the water volume of inlet water, outlet water, backwash water and return water in real time, and issues early warning in time to accurately control the supply-supply difference rate. At the same time, it adjusts the liquid level of the clear water reservoir to maintain it at a reasonable high level, thereby effectively reducing the power consumption of the water delivery unit during the water supply process.

[0085] Intelligent monitoring module, which comprehensively collects the operating status and fault signals of all process equipment in the water plant, including water pumps, blowers, valves, mixers, etc. At the same time, the module is also responsible for summarizing the data of various testing instruments such as flow meters, liquid level meters, pressure gauges, turbidity analyzers, residual chlorine monitors, pH meters, as well as the information of all intelligent instruments such as power monitoring instruments and temperature inspection systems, thereby realizing centralized and efficient monitoring of the production data of the entire water plant.

[0086] Equipment management module: This module manages the entire life cycle of the plant's process equipment, including basic equipment ledger information, equipment operation, equipment alarms, equipment repair and maintenance plans, equipment repair and maintenance lists, and equipment residual values, etc., providing a decision-making basis for the water plant's water production cost and equipment expenses.

[0087] In summary, the embodiment of the present invention provides a control method and control system for a multi-agent water plant process system. Fig.10As shown, first, independent process simulation is performed based on the water environment and water management objectives of the water plant to obtain the optimal control solution set of the single agent; secondly, combined with the competition and game relationship between each single agent, through multi-objective iterative optimization, the multi-agent collaborative control strategy solution set is obtained; then, combined with the single cost status and the comprehensive effect of the solution, the multi-agent collaborative control strategy that meets the control objective function is obtained; finally, according to the real-time change information of the water plant process system such as water environment, objectives, and costs, the multi-agent collaborative control strategy is dynamically adjusted through iteration to achieve the dynamic response of the control system to real-time changes. The present invention provides a new set of dynamic optimal collaborative control solutions for the water plant process system, which manages the competition and cooperation relationship between single agents through an efficient coordination mechanism to ensure that each single agent will not damage the overall harmony and water production efficiency of the water plant process system while pursuing its own optimization goals. At the same time, the present invention can also adaptively adjust the collaborative control strategy of the water plant process system according to the operating data of the water plant process system and the external environment change information, and realize more precise and dynamic control by continuously optimizing the control strategy process. It not only significantly improves the stability and intelligence level of the water plant process system operation, but also effectively reduces the energy consumption of the water plant process system operation and improves the resource utilization efficiency, so as to ensure that the indicators of water quality and water quantity can be continuously optimized.

[0088] Since the system / device described in the above embodiments of the present invention is a system / device used to implement the method of the above embodiments of the present invention, a person skilled in the art can understand the specific structure and deformation of the system / device based on the method described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the method of the above embodiments of the present invention belong to the scope of protection of the present invention.

[0089] It should be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0090] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0091] It should be noted that in the description of the present invention, the word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The present invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. The use of the words first, second, third, etc., is only for convenience of expression and does not indicate any order. These words may be understood as part of the name of the component.

[0092] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0093] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments after obtaining the basic inventive concepts.

[0094] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the invention.

Claims

1. A control method for a multi-agent water plant process system, characterized in that: include: Taking the minimization of production cost and optimization of water quality of the water plant process system as the dual control objectives, a control objective function for maximizing the global benefits of the water plant process system is constructed; According to the preset constraints, each single agent of the water plant process system is independently simulated to obtain the optimal control solution for each single agent; The optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set, and the multi-agent collaborative control scheme set is subjected to multi-objective iterative optimization including exchange optimality, production optimality and hybrid optimality according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function; According to the real-time change information of the water plant process system, the multi-agent collaborative control strategy is dynamically adjusted, and the adjusted optimal multi-agent collaborative control strategy is deployed to the water plant process system.

2. The method according to claim 1, characterized in that Taking the minimization of production cost and optimization of water quality of the water plant process system as the dual control objectives, the control objective function for maximizing the global benefit of the water plant process system is constructed, including: Obtain the production cost information and water quality information of the water plant process system. The production cost information includes the water production process cost and the water supply process cost; According to the water supply quality information and the preset water supply quality target, the deviation between the water supply quality and the water supply quality target is obtained; Based on the preset collaborative reward function and interference penalty function between single agents, combined with the production cost information and the deviation between the actual water quality and the water quality target, a control objective function for maximizing the global benefit of the water plant process system is constructed with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system. Among them, the control objective function is: ； In the formula, k 1 is the weight of water quality, k 2 is the weight of water production cost, k 3 is the weight of water supply cost, k 4 is the weight of the collaborative reward function, k 5 is the weight of the interference penalty function, is the deviation between the actual water quality and the water quality target. V punit The cost of water production process, FT in is the water inlet flow rate, V sunit The cost of water supply process, PT is the actual water supply pressure, FT out is the water supply flow rate, R 1 is the collaborative reward function between single agents, R 2 is the interference penalty function between single agents.

3. The method according to claim 1, characterized in that According to the preset constraints, each single agent of the water plant process system is independently simulated to obtain the optimal control solution for each single agent, including: According to the process type of each single agent in the water plant process system, the dual control objectives are disassembled to obtain the dual control sub-objectives of minimizing the production cost of each single agent and optimizing the output result; Traverse all dual control sub-goals, and obtain the constraint conditions for each single agent to perform process simulation based on the influencing factors of the single agent operation; Conduct independent process simulation for each single agent of the water plant process system according to the constraints, analyze the simulation output results and process costs, and obtain the optimal control solution for each single agent; The constraint condition expression is: ； In the formula, For a single agent i The minimum dosage of the agent, For a single agent i The dosage of the agent, For a single agent i The maximum dosage of the agent, For a single agent i The minimum liquid level, Represents a single agent i The liquid level, Represents a single agent i The maximum liquid level, For a single agent i Medium Equipment j The number of runs, For a single agent i Medium Equipment j The maximum number of runs, For a single agent i Medium Equipment a and equipment b Only one can be run at a time.

4. The method according to claim 1, characterized in that The optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set, and the multi-agent collaborative control scheme set is subjected to multi-objective iterative optimization including exchange optimality, production optimality and hybrid optimality according to the control objective function. The multi-agent collaborative control strategy that satisfies the control objective function includes: According to the acquired single-agent process type, the optimal control schemes are grouped and integrated to form at least one multi-agent collaborative control scheme set; According to the preset reward and punishment function between single agents, the multi-agent collaborative control solution set is optimized by multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization to obtain the multi-agent collaborative control strategy solution set; According to the control objective function of maximizing the global benefit of the water plant process system, the multi-agent collaborative control strategy solution set is converged to obtain a multi-agent collaborative control strategy that meets the control objective function; Among them, the reward and punishment functions include: collaboration reward function and interference penalty function; The collaborative reward function is: ； In the formula, λ is the weight of the efficient collaboration reward, δ the weight of the reward for inefficient collaboration; The interference penalty function is: R 2=－ μ ； In the formula, μ The intensity of the interference penalty.

5. The method according to claim 4, characterized in that According to the preset reward and punishment function between single agents, the multi-agent collaborative control solution set is optimized by multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization, and the multi-agent collaborative control strategy solution set is obtained, including: According to the preset reward and punishment function between single agents, the reward and punishment weights between single agents are obtained; Perform multi-objective coordinated optimization on the multi-agent collaborative control solution set, including exchange optimization, production optimization and hybrid optimization, to obtain the initial multi-agent collaborative control strategy solution set; Through the genetic multi-objective optimization algorithm and combined with the reward and punishment weights, the initial multi-agent collaborative control strategy solution set is iteratively updated to obtain the multi-agent collaborative control strategy solution set.

6. The method according to claim 4, characterized in that According to the control objective function of maximizing the global benefit of the water plant process system, the multi-agent collaborative control strategy solution set is converged to obtain the multi-agent collaborative control strategy that meets the control objective function, including: Traverse each multi-agent collaborative control strategy in the multi-agent collaborative control strategy solution set, and select and obtain a multi-agent collaborative control strategy that meets the set minimum water quality target; According to the control objective function of maximizing the global benefit of the water plant process system, the global benefit data of the water plant process system of each multi-agent collaborative control strategy after the minority selection is obtained; The multi-agent collaborative control strategy with the largest global benefit data of the water plant process system is determined as the multi-agent collaborative control strategy that meets the control objective function.

7. The method according to claim 1, characterized in that According to the real-time change information of the water plant process system, the multi-agent collaborative control strategy is dynamically adjusted, and the optimal multi-agent collaborative control strategy obtained by adjustment is deployed to the water plant process system, including: Obtain real-time change information of the water plant process system, including water plant environment change information, water supply target change information and production cost change information; According to the real-time change information, the multi-agent collaborative control strategy is dynamically adjusted to obtain the optimal multi-agent collaborative control strategy for the water plant process system; The optimal multi-agent collaborative control strategy is deployed to the water plant process system to make the operating conditions of the water plant process system optimal.

8. A control system for a multi-agent water plant process system, characterized in that: include: The control objective function construction module is used to construct the control objective function of maximizing the global benefit of the water plant process system with the dual control objectives of minimizing the production cost and optimizing the water quality of the water plant process system; The single-agent control optimization module is used to perform independent process simulation on each single-agent of the water plant process system according to preset constraints to obtain the optimal control solution for each single-agent; The multi-agent control optimization module is used to group and integrate the optimal control schemes to form at least one multi-agent collaborative control scheme set, and perform multi-objective iterative optimization including exchange optimization, production optimization and hybrid optimization on the multi-agent collaborative control scheme set according to the control objective function to obtain a multi-agent collaborative control strategy that meets the control objective function; The control strategy dynamic adjustment module is used to dynamically adjust the multi-agent collaborative control strategy based on the real-time change information of the water plant process system, and deploy the adjusted optimal multi-agent collaborative control strategy to the water plant process system.

9. A multi-agent water plant process system, characterized in that: include: A water plant process brain module, used to execute the control method steps of a multi-agent water plant process system as described in any one of claims 1 to 7; The single intelligent agent module of the water plant is connected to the process brain module of the water plant, and is used to execute the water production process of the water plant when receiving the control instructions of the process brain module of the water plant.

10. The multi-agent water plant process system according to claim 9, characterized in that: The single intelligent module of the water plant includes: intelligent alum addition module, intelligent chlorination module, intelligent backwash module, intelligent sludge discharge module, intelligent pump distribution module, intelligent water balance module, intelligent monitoring module and equipment management module.

Images