A method for cooling station load distribution based on integrated optimization algorithm

Abstract

The present invention discloses a cold station load distribution method based on an integrated optimization algorithm, and relates to the technical field of cold station load distribution. The cold station load distribution method based on the integrated optimization algorithm includes the following steps: S1, establishing a cold station model; S2, constructing an optimization model for cold station load distribution; S3, solving the optimization model with an integrated optimization algorithm; S4, applying an optimal load distribution scheme. The cold station load distribution method based on the integrated optimization algorithm combines and integrates a variety of optimization algorithms, utilizes the advantages of different algorithms, and realizes the advantages and disadvantages on the basis of information sharing. While improving the solution efficiency, it enhances the global optimization ability of the algorithm, overcomes the limitations of a single optimization algorithm, can effectively improve the efficiency and accuracy of load distribution, reduce the energy consumption of cold stations, and has a good energy-saving effect.

Description

Technical Field

[0001] The invention relates to the technical field of refrigeration station load distribution, and in particular to a refrigeration station load distribution method based on an integrated optimization algorithm. Background Art

[0002] With the continuous acceleration of urbanization, centralized cooling systems have been widely used in urban construction. As an important part of modern urban construction, centralized cooling systems can significantly improve energy utilization efficiency through integrated management and operation. Centralized cooling systems can make full use of energy and reduce energy consumption through centralized cooling, so they have good energy-saving effects and economic benefits. In centralized cooling systems, reasonable distribution of cooling station loads is of great significance to reducing the energy consumption of cooling systems and improving economic benefits. In such systems, reasonable distribution of the loads of various cooling stations is not only related to the operating efficiency of the entire system, but also directly affects the overall economic cost. Due to differences in the demand for cooling in different regions and at different times, how to dynamically adjust the working status of each cooling station according to real-time demand has become the key to ensuring efficient operation of the system.

[0003] At present, the load distribution of cold stations mainly adopts traditional optimization methods, such as linear programming, dynamic programming, etc. Although traditional optimization methods can solve the problem to a certain extent, they often need to be adjusted for specific problems, and when faced with complex and changeable actual scenarios, they may not be able to quickly find the global optimal solution. These methods are generally targeted at specific problems and lack versatility and robustness. In addition, traditional optimization algorithms are prone to fall into local optimality and have slow solution speeds, making it difficult to meet the needs of actual engineering projects.

[0004] Therefore, in view of this, the existing structure and deficiencies are studied and improved, and a cold station load distribution method based on integrated optimization algorithm is proposed. Summary of the invention

[0005] In view of the deficiencies of the prior art, the present invention provides a method for distributing cold station loads based on an integrated optimization algorithm, which solves the problems raised in the above-mentioned background technology.

[0006] To achieve the above objectives, the present invention is implemented by the following technical scheme: a cooling station load distribution method based on an integrated optimization algorithm, the cooling station load distribution method based on an integrated optimization algorithm comprises the following steps:

[0007] S1. Establish cold station model:

[0008] By installing sensors on various subsystems and equipment in the cold station, key parameters such as the inlet and outlet water temperature, chilled water flow, and cooling capacity of the chiller are collected in real time, and the data is transmitted to the database of the cold station control system for storage. On this basis, the characteristic curve of the chiller, the thermodynamic model of the cooling tower, and the water pump energy consumption model are integrated to establish a mathematical model that can accurately reflect the actual operating status of the cold station;

[0009] The mathematical model that reflects the actual operating status of the cold station is a complex dynamic system, which is difficult to be fully described by a single mathematical model. However, different mathematical models can be used to describe different aspects, such as:

[0010] Refrigerant flow model:

[0011] For the refrigerant flow, the following mathematical model can be used:

[0012] Q_ref=C_ref*(T_ref-T_out);

[0013] in:

[0014] Q_ref: refrigerant flow rate (kg / s);

[0015] C_ref: refrigerant flow coefficient (related to refrigerant type, pipe diameter, etc.);

[0016] T_ref: refrigerant inlet and outlet temperature (℃);

[0017] T_out: refrigerant outlet temperature (℃);

[0018] Refrigerant heat balance model:

[0019] The heat balance of the refrigerant can be described by the law of conservation of energy:

[0020] Q_in-Q_out=m_ref*(h_in-h_out);

[0021] in:

[0022] Q_in: refrigerant inlet and outlet heat (kJ / s);

[0023] Q_out: refrigerant outlet heat (kJ / s);

[0024] m_ref: refrigerant flow rate (kg / s);

[0025] h_in: refrigerant inlet and outlet enthalpy (kJ / kg);

[0026] h_out: refrigerant outlet enthalpy (kJ / kg);

[0027] The refrigerant enthalpy can be estimated using the following formula:

[0028] h_ref=c_p*T_ref+h_ref0+(T_ref-T_ref0)*(c_p / 1000);

[0029] in:

[0030] h_ref: refrigerant enthalpy (kJ / kg);

[0031] c_p: specific heat capacity of refrigerant (kJ / kg·K);

[0032] h_ref0: reference value of refrigerant enthalpy (kJ / kg);

[0033] T_ref: refrigerant temperature (℃);

[0034] T_ref0: refrigerant temperature reference value (℃);

[0035] Freezing power model:

[0036] Q_evap = m_ref*(h1-h4);

[0037] Where, Q_evap represents the refrigeration power, m_ref represents the mass flow rate of the condensing medium, h1 and h4 represent the specific enthalpy of the condensing medium at the inlet and outlet of the evaporator respectively;

[0038] Condensing power model:

[0039] Q_cond = m_ref*(h2-h3);

[0040] Where, Q_cond represents the condensing power, h2 and h3 represent the specific enthalpy of the condensing medium at the inlet and outlet of the condenser respectively;

[0041] Pressure loss model:

[0042]

[0043] Where Δp represents the pressure loss, ζ represents the local resistance coefficient, ρ_ref represents the density of the condensing medium, v represents the velocity of the condensing medium, and the sub-foots comp, cond and evap represent the compressor, condenser and evaporator respectively;

[0044] Energy efficiency ratio model:

[0045] COP = Q_evap / W_comp;

[0046] Where, COP represents energy efficiency ratio, W_comp represents the power input of the compressor;

[0047] Heat exchange efficiency model of condenser and evaporator:

[0048]

[0049] Among them, ε represents the heat exchange efficiency, T represents the temperature, and the sub-pins cond and evap represent the condenser and evaporator respectively;

[0050] S2. Constructing an optimization model for cooling station load distribution:

[0051] Taking minimizing the total energy consumption of the cooling station as the objective function, combined with the operating condition constraints of the chiller, the balance of chilled water flow and other conditions, an optimization model for the load distribution of the cooling station is constructed. The optimization model uses the load distribution ratio of each cooling station as the decision variable, and quantifies the objective function value of each scheme through the energy consumption calculation formula.

[0052] In the energy consumption calculation formula quantification scheme, the objective function value is usually used to evaluate the energy efficiency and energy-saving effect of different schemes. The objective function is as follows:

[0053] Minimize total energy consumption:

[0054] E_total = ∑(E_i);

[0055] Among them, E_total represents the total energy consumption, and E_i represents the energy consumption of the i-th device or system;

[0056] Minimize energy costs:

[0057] C_total = ∑(C_i*E_i);

[0058] Among them, C_total represents the total energy cost, C_i represents the energy unit price of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0059] Minimizing Carbon Emissions:

[0060] GHG_total = ∑(GHG_i*E_i);

[0061] Among them, GHG_total represents the total carbon emissions, GHG_i represents the carbon emission coefficient of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0062] Maximize energy efficiency:

[0063] η_total=∑(η_i*E_i) / ∑(E_i);

[0064] Among them, η_total represents the total energy efficiency, η_i represents the energy efficiency of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0065] Maximize energy saving effect:

[0066] S_total = ∑(S_i*E_i);

[0067] Among them, S_total represents the total energy saving effect, S_i represents the energy saving effect of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0068] In practical applications, it may be necessary to combine multiple objective functions to evaluate the advantages and disadvantages of different solutions and make comprehensive optimization and decisions. It should be noted that when calculating the objective function value, various constraints and restrictions need to be considered, such as the power limit of the equipment, the reliability of energy supply, the influence of environmental factors, etc.

[0069] S3. Integrated optimization algorithm to solve the optimization model:

[0070] The core idea of ​​the integrated optimization algorithm is to organically combine multiple heuristic algorithms and achieve collaborative optimization between algorithms through an information exchange mechanism. Specifically, the integrated optimization algorithm includes the following sub-steps:

[0071] A: parameter initialization;

[0072] B: Parallel optimization;

[0073] C: Solution evaluation and information exchange;

[0074] D: Algorithm update;

[0075] E: Iterative solution;

[0076] F: result output;

[0077] S4. Application of optimal load distribution scheme:

[0078] According to the optimal load distribution ratio, the operating parameter setting values ​​of each cold station and its internal chillers are calculated, such as the start and stop status of the chiller, the chilled water outlet temperature, the chilled water flow rate, etc. These control instructions are sent to the on-site controller of each cold station through the data bus. The controller adjusts the operating status of the chiller according to the instructions to make the actual operating conditions close to the optimal solution. At the same time, the cold station control system also needs to periodically trigger the load distribution optimization program, obtain the operating condition changes of the cold station in real time, and dynamically update the optimization results to ensure that the cold station can achieve optimal energy consumption operation under different load demands.

[0079] Furthermore, in step S2, according to the actual operating conditions of the cold station, constraints such as the range of the number of operating chillers and the upper and lower limits of the chilled water outlet temperature are set to ensure that the optimization result meets the feasibility requirements.

[0080] Furthermore, in step S3, the specific contents of parameter initialization are as follows:

[0081] The common parameters of particle swarm optimization algorithm, genetic algorithm and simulated annealing algorithm are uniformly initialized, including the initial solution set, fitness evaluation function and termination condition.

[0082] Furthermore, in step S3, the specific contents of parallel optimization are as follows:

[0083] Particle swarm, genetic and simulated annealing algorithms are optimized in parallel in their respective search spaces, searching for the optimal load distribution solution from different angles. The particle swarm algorithm realizes the iteration of the solution by updating the position and velocity of particles, the genetic algorithm realizes the evolution of the population through selection, crossover and mutation operations, and the simulated annealing algorithm controls the probability of accepting a new solution through the temperature parameter.

[0084] Furthermore, in step S3, the specific contents of solution evaluation and information exchange are as follows:

[0085] The current solutions of the three algorithms are evaluated uniformly on a regular basis, and the objective function values ​​of each solution are calculated. Based on the evaluation results, the high-quality solutions of each algorithm are identified, and information is exchanged through a shared memory mechanism. The particle swarm algorithm shares its global optimal solution with the genetic algorithm and joins the population as a new high-quality individual. The genetic algorithm shares the individual with the highest fitness in its population with the simulated annealing algorithm as the new current solution. The simulated annealing algorithm shares its accepted new solution with the particle swarm algorithm to help particles escape from the local optimal area.

[0086] Furthermore, in step S3, the specific contents of the algorithm update are as follows:

[0087] Based on the exchanged information, each algorithm adaptively updates its own key parameters. For example, the particle swarm algorithm adjusts the speed of particles according to the shared high-quality solutions, the genetic algorithm dynamically adjusts the crossover and mutation probabilities according to the introduced external individuals, and the simulated annealing algorithm adjusts the temperature decay rate according to the energy consumption level of the shared solutions.

[0088] Furthermore, in step S3, the specific contents of iterative solution are as follows:

[0089] The three updated algorithms enter the parallel optimization stage again and repeat steps BD until the maximum number of iterations or the accuracy requirement of the solution is met. In each round of iteration, the algorithms continuously exchange information.

[0090] Furthermore, in step S3, the specific content of the result output is as follows:

[0091] When the termination condition is met, the load distribution scheme with the lowest total energy consumption is selected from the optimization results of the three algorithms as the final global optimal solution. The optimal solution is decoded into the load distribution ratio of each cooling station and transmitted to the next control module.

[0092] Furthermore, the cooling station load distribution method based on the integrated optimization algorithm is applied to the technical field of cooling station load distribution.

[0093] The present invention provides a cooling station load distribution method based on an integrated optimization algorithm, which has the following beneficial effects:

[0094] This cold station load distribution method based on integrated optimization algorithm combines and integrates multiple optimization algorithms, utilizes the advantages of different algorithms, and realizes the best of both worlds on the basis of information sharing. While improving the solution efficiency, it enhances the global optimization ability of the algorithm, overcomes the limitations of a single optimization algorithm, and can effectively improve the efficiency and accuracy of load distribution, reduce the energy consumption of cold stations, and has a good energy-saving effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] Figure 1 It is a schematic diagram of the overall process of a cold station load distribution method based on an integrated optimization algorithm of the present invention;

[0096] Figure 2 It is a schematic diagram of an integrated optimization algorithm flow of a cold station load distribution method based on an integrated optimization algorithm of the present invention;

[0097] Figure 3 A schematic diagram of system data flow of a cold station load distribution method based on an integrated optimization algorithm of the present invention;

[0098] Figure 4 The present invention is a schematic diagram of an algorithm information exchange process of a cold station load distribution method based on an integrated optimization algorithm. DETAILED DESCRIPTION

[0099] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0100] like Figure 1-Figure 4 As shown, the present invention provides a technical solution: a cold station load distribution method based on an integrated optimization algorithm, the cold station load distribution method based on an integrated optimization algorithm comprises the following steps:

[0101] S1. Establish cold station model:

[0102] By installing sensors on various subsystems and equipment in the cold station, the key parameters of the chiller, such as the inlet and outlet water temperature, chilled water flow, and cooling capacity, are collected in real time, and the data is transmitted to the database of the cold station control system for storage. On this basis, the characteristic curve of the chiller, the thermodynamic model of the cooling tower, the water pump energy consumption model and other models are integrated to establish a mathematical model that can accurately reflect the actual operating status of the cold station;

[0103] The mathematical model that reflects the actual operating status of the cold station is a complex dynamic system, which is difficult to be fully described by a single mathematical model. However, different mathematical models can be used to describe different aspects, such as:

[0104] Refrigerant flow model:

[0105] For the refrigerant flow rate, the following mathematical model can be used: Q_ref = C_ref*(T_ref-T_out);

[0106] in:

[0107] Q_ref: refrigerant flow rate (kg / s);

[0108] C_ref: refrigerant flow coefficient (related to refrigerant type, pipe diameter, etc.); T_ref: refrigerant inlet and outlet temperature (℃);

[0109] T_out: refrigerant outlet temperature (℃);

[0110] Refrigerant heat balance model:

[0111] The heat balance of the refrigerant can be described by the law of conservation of energy: Q_in-Q_out=m_ref*(h_in-h_out);

[0112] in:

[0113] Q_in: refrigerant inlet and outlet heat (kJ / s);

[0114] Q_out: refrigerant outlet heat (kJ / s);

[0115] m_ref: refrigerant flow rate (kg / s);

[0116] h_in: refrigerant inlet and outlet enthalpy (kJ / kg);

[0117] h_out: refrigerant outlet enthalpy (kJ / kg);

[0118] The refrigerant enthalpy can be estimated using the following formula: h_ref = c_p*T_ref+h_ref0+(T_ref-T_ref0)*(c_p / 1000); where:

[0119] h_ref: refrigerant enthalpy (kJ / kg);

[0120] c_p: specific heat capacity of refrigerant (kJ / kg·K);

[0121] h_ref0: reference value of refrigerant enthalpy (kJ / kg);

[0122] T_ref: refrigerant temperature (℃);

[0123] T_ref0: refrigerant temperature reference value (℃);

[0124] Freezing power model:

[0125] Q_evap = m_ref*(h1-h4);

[0126] Where, Q_evap represents the refrigeration power, m_ref represents the mass flow rate of the condensing medium, h1 and h4 represent the specific enthalpy of the condensing medium at the inlet and outlet of the evaporator respectively;

[0127] Condensing power model:

[0128] Q_cond = m_ref*(h2-h3);

[0129] Where, Q_cond represents the condensing power, h2 and h3 represent the specific enthalpy of the condensing medium at the inlet and outlet of the condenser respectively;

[0130] Pressure loss model:

[0131]

[0132] Where Δp represents the pressure loss, ζ represents the local resistance coefficient, ρ_ref represents the density of the condensing medium, v represents the velocity of the condensing medium, and the sub-foots comp, cond and evap represent the compressor, condenser and evaporator respectively;

[0133] Energy efficiency ratio model:

[0134] COP = Q_evap / W_comp;

[0135] Where, COP represents energy efficiency ratio, W_comp represents the power input of the compressor;

[0136] Heat exchange efficiency model of condenser and evaporator:

[0137]

[0138] Among them, ε represents the heat exchange efficiency, T represents the temperature, and the sub-pins cond and evap represent the condenser and evaporator respectively;

[0139] S2. Constructing an optimization model for cooling station load distribution:

[0140] Taking minimizing the total energy consumption of the cooling station as the objective function, combined with the operating condition constraints of the chiller, the balance of chilled water flow and other conditions, an optimization model for the load distribution of the cooling station is constructed. The optimization model uses the load distribution ratio of each cooling station as the decision variable, and quantifies the objective function value of each scheme through the energy consumption calculation formula.

[0141] In the energy consumption calculation formula quantification scheme, the objective function value is usually used to evaluate the energy efficiency and energy-saving effect of different schemes. The objective function is as follows:

[0142] Minimize total energy consumption:

[0143] E_total = ∑(E_i);

[0144] Among them, E_total represents the total energy consumption, and E_i represents the energy consumption of the i-th device or system;

[0145] Minimize energy costs:

[0146] C_total = ∑(C_i*E_i);

[0147] Among them, C_total represents the total energy cost, C_i represents the energy unit price of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0148] Minimizing Carbon Emissions:

[0149] GHG_total = ∑(GHG_i*E_i);

[0150] Among them, GHG_total represents the total carbon emissions, GHG_i represents the carbon emission coefficient of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0151] Maximize energy efficiency:

[0152] η_total=∑(η_i*E_i) / ∑(E_i);

[0153] Among them, η_total represents the total energy efficiency, η_i represents the energy efficiency of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0154] Maximize energy saving effect:

[0155] S_total = ∑(S_i*E_i);

[0156] Among them, S_total represents the total energy saving effect, S_i represents the energy saving effect of the i-th device or system, and E_i represents the energy consumption of the i-th device or system;

[0157] In practical applications, it may be necessary to combine multiple objective functions to evaluate the advantages and disadvantages of different solutions and make comprehensive optimization and decisions. It should be noted that when calculating the objective function value, various constraints and restrictions need to be considered, such as the power limit of the equipment, the reliability of energy supply, the influence of environmental factors, etc.

[0158] S3. Integrated optimization algorithm to solve the optimization model:

[0159] The core idea of ​​the integrated optimization algorithm is to organically combine multiple heuristic algorithms and achieve collaborative optimization between algorithms through an information exchange mechanism. Specifically, the integrated optimization algorithm includes the following sub-steps:

[0160] A: parameter initialization;

[0161] The specific contents of parameter initialization are as follows:

[0162] The common parameters of particle swarm optimization algorithm, genetic algorithm and simulated annealing algorithm are uniformly initialized, including the initial solution set, fitness evaluation function and termination condition. This unified parameter setting is conducive to fair competition and information sharing among algorithms.

[0163] B: Parallel optimization;

[0164] The specific contents of parallel optimization are as follows:

[0165] Particle swarm, genetic and simulated annealing algorithms are optimized in parallel in their respective search spaces, searching for the optimal load distribution solution from different angles. The particle swarm algorithm realizes the iteration of the solution by updating the position and velocity of particles, the genetic algorithm realizes the evolution of the population through selection, crossover and mutation operations, and the simulated annealing algorithm controls the probability of accepting a new solution through the temperature parameter.

[0166] C: Solution evaluation and information exchange;

[0167] The specific contents of the evaluation and information exchange are as follows:

[0168] Regularly conduct a unified fitness evaluation on the current solutions of the three algorithms, calculate the objective function value of each solution, identify the high-quality solutions of each algorithm based on the evaluation results, and exchange information through a shared memory mechanism. Among them, the particle swarm algorithm shares its global optimal solution with the genetic algorithm and joins the population as a new high-quality individual. The genetic algorithm shares the individual with the highest fitness in its population with the simulated annealing algorithm as the new current solution. The simulated annealing algorithm shares its accepted new solution with the particle swarm algorithm to help particles jump out of the local optimal area.

[0169] D: Algorithm update;

[0170] The specific contents of the algorithm update are as follows:

[0171] Based on the exchanged information, each algorithm adaptively updates its key parameters. For example, the particle swarm algorithm adjusts the speed of particles according to the shared high-quality solutions, the genetic algorithm dynamically adjusts the crossover and mutation probabilities according to the external individuals introduced, and the simulated annealing algorithm adjusts the temperature decay rate according to the energy consumption level of the shared solutions. This parameter adaptive update effectively balances the local search and global exploration capabilities of the algorithm.

[0172] E: Iterative solution;

[0173] The specific contents of iterative solution are as follows:

[0174] The three updated algorithms enter the parallel optimization phase again and repeat steps BD until the maximum number of iterations or the accuracy requirements of the solution are met. In each round of iteration, the algorithms continuously exchange information, promote each other, and search together towards the direction of the global optimal solution;

[0175] F: result output;

[0176] The output results are as follows:

[0177] When the termination condition is met, the load distribution scheme with the lowest total energy consumption is selected from the optimization results of the three algorithms as the final global optimal solution. The optimal solution is decoded into the load distribution ratio of each cooling station and transmitted to the next control module.

[0178] S4. Application of optimal load distribution scheme:

[0179] According to the optimal load distribution ratio, the operating parameter setting values ​​of each cold station and its internal chillers are calculated, such as the start and stop status of the chiller, the chilled water outlet temperature, the chilled water flow rate, etc. These control instructions are sent to the on-site controller of each cold station through the data bus. The controller adjusts the operating status of the chiller according to the instructions to make the actual operating conditions close to the optimal solution. At the same time, the cold station control system also needs to periodically trigger the load distribution optimization program, obtain the operating condition changes of the cold station in real time, and dynamically update the optimization results to ensure that the cold station can achieve optimal energy consumption operation under different load demands.

[0180] The cooling station load distribution method based on the integrated optimization algorithm is applied to the technical field of cooling station load distribution.

[0181] The cold station load distribution method based on the integrated optimization algorithm is applied to a computer running on a cold station control system, the computer including a data acquisition module, a model building module, an optimization model building module, an integrated optimization solution module and a control instruction generation module;

[0182] Data acquisition module: including sensors distributed in various subsystems of the cold station, used to extract the real-time operating condition data and equipment parameters of the cold station from the database of the cold station control system, such as the inlet and outlet water temperature, chilled water flow, refrigeration capacity, etc. of the chiller, and perform necessary data preprocessing;

[0183] Model building module: extract historical data from the database, analyze and process it, and then build a mathematical model of the cooling station system. According to the optimization goal, build an optimization model for load distribution, which is used to build a mathematical model of the cooling station system based on the collected data. The model comprehensively considers factors such as chiller characteristics, cooling tower performance, and water pump characteristics, and can accurately describe the actual operating status of the cooling station.

[0184] Optimization model building module: used to minimize the energy consumption of the cold station as the optimization goal, combined with the operating conditions constraints of the chiller, the balance of chilled water flow and other conditions, to build an optimization model for the load distribution of the cold station. The model can flexibly set optimization variables and constraints according to actual needs;

[0185] Integrated optimization solution module: It has multiple built-in heuristic algorithms. It solves the load distribution model through integrated optimization to obtain the global optimal load distribution ratio. This module is the core of the optimization program. It calls the integrated optimization algorithm to solve the above optimization model. Through heuristic rules and iterative search strategies, it realizes the collaborative optimization of particle swarm algorithm, genetic algorithm and simulated annealing algorithm, and quickly searches for the global optimal load distribution solution.

[0186] Control instruction generation module: converts the calculation results of integrated optimization into cold station control instructions, mainly including the number of operating units of each chiller, the set value of chilled water outlet temperature, etc. The generated control instructions are sent to the field controller through the data bus to guide the energy-saving operation of the chiller;

[0187] Among them, the particle swarm algorithm steps include:

[0188] 1. Initialize the particle swarm: randomly generate a group of particles, each particle represents a candidate solution, including position (representing decision variables) and velocity;

[0189] 2. Initialize the best position: set a variable p for each particle to record its best position best , initially p best Set to current position;

[0190] 3. Calculate the global optimal position: Find the particle with the current best position in the particle swarm and set its position to the global optimal position g best ;

[0191] 4. Update particle speed and position: according to p best , g bestand a set of randomly generated acceleration factors to update the velocity and position of each particle using the following formula:

[0192] Speed ​​Update:

[0193] v i (t+1)=w×v i (t)+c1×r1×(p best_i -x i (t))+c2×r2×(g best -x i (t)) (1)

[0194] Location Updates:

[0195] x_i(t+1)=x_i(t)+v_i(t+1) (2)

[0196] Where x_i(t) and v_i(t) represent the velocity and position of particle i at the tth iteration, w is the inertia weight, c1 and c2 are acceleration constants, and r1 and r2 are random numbers;

[0197] 5. Evaluate particle position: Use the objective function to evaluate the position of each particle and update p best and g best , if the current position is better, update p best ; If the current position is better than the global best position, update g best ;

[0198] 6. Check the termination condition: If the maximum number of iterations is reached or other termination conditions (such as convergence criteria) are met, the algorithm is terminated and the current best position is returned as the solution to the optimization problem. Otherwise, return to step 4 to continue iterating.

[0199] The genetic algorithm steps include:

[0200] 1. Initialize the chromosome group: randomly generate a group of chromosomes, each chromosome represents a candidate solution, and consists of a string of binary codes or real number codes;

[0201] 2. Evaluate chromosomes: Use the objective function to evaluate the fitness or fitness value of each chromosome. The fitness indicates the quality of the solution encoded by the chromosome.

[0202] 3. Selection operation: According to the chromosome fitness, use the selection operation to select a part of chromosomes from the current chromosome group as parents to participate in the generation of the next generation. The higher the fitness, the greater the probability of being selected;

[0203] IV. Crossover operation: Perform crossover operation on the selected parental chromosomes, exchange or recombine a part of the chromosomes to generate new chromosomes as offspring. Common crossover operations include single-point crossover, multi-point crossover, and uniform crossover, etc.;

[0204] V. Mutation operation: Perform mutation operation on the offspring chromosomes, randomly change a part of the genes on the chromosomes to increase diversity and avoid falling into local optimal solutions. Common mutation operations include bit flipping, gene exchange, etc.;

[0205] VI. Replace chromosomes: Replace a part of the current chromosome population with the newly generated offspring chromosomes to form a new chromosome population;

[0206] VII. Check termination conditions: If the maximum number of iterations is reached or other termination conditions (such as convergence criteria) are met, terminate the algorithm and return the current best chromosome as the solution to the optimization problem. Otherwise, return to Step II to continue the iteration;

[0207] The steps of the simulated annealing algorithm include:

[0208] I. Initialization: Randomly generate an initial solution and set the initial temperature T;

[0209] II. Evaluate the current solution: Use the objective function to evaluate the fitness or objective value of the current solution;

[0210] III. Generate a new solution: Randomly generate a new solution based on the current solution;

[0211] IV. Calculate the acceptance probability: Calculate the probability p of accepting the new solution according to the current temperature T and the fitness difference between the current solution and the new solution. The acceptance probability is usually given by the following formula:

[0212] p = exp((fitness of current solution - fitness of new solution) / T)

[0213] where exp represents the natural exponential function. If the fitness of the new solution is higher than that of the current solution, then p = 1, indicating that the new solution is definitely accepted;

[0214] V. Compare the new solution: According to the acceptance probability p, compare whether the new solution is accepted. Generate a random number r between 0 and 1. If r < p, accept the new solution; otherwise, keep the current solution;

[0215] VI. Decrease the temperature: Decrease the temperature T according to the annealing schedule. Common annealing schedules include geometric annealing (T_new = α * T_old, 0 < α < 1) and exponential annealing (T_new = T_old / number of iterations), etc.;

[0216] 7. Check the termination conditions: If the maximum number of iterations is reached or other termination conditions (such as convergence criteria) are met, the algorithm is terminated and the current best solution is returned as the solution to the optimization problem. Otherwise, return to step 3 to continue iterating.

[0217] In summary, when using the cold station load distribution method based on the integrated optimization algorithm, the operating condition data and equipment parameters of the cold station system are first obtained, and a mathematical model of the cold station system is established. Taking minimizing the energy consumption of the cold station as the objective function, the optimization model of the cold station load distribution is constructed in combination with equipment operation constraints, load balance and other conditions;

[0218] The optimization model is solved by using an integrated optimization algorithm to obtain the optimal load distribution ratio of each cooling station. The integrated optimization algorithm includes the following steps:

[0219] Initialization: uniformly initialize the parameters of particle swarm optimization algorithm, genetic algorithm, and simulated annealing algorithm, including shared initial solutions, unified termination conditions, etc.;

[0220] Parallel optimization: Particle swarm optimization, genetic algorithm, and simulated annealing algorithm are optimized in parallel in their respective search spaces, while searching for the optimal load distribution solution;

[0221] Solution evaluation: Perform a unified fitness evaluation on the load distribution solutions obtained by the three algorithms and calculate the cooling station energy consumption of each solution;

[0222] Information exchange and update: Based on the evaluation results, three algorithms are used to exchange and share information:

[0223] The particle swarm algorithm shares its global optimal solution with the genetic algorithm and simulated annealing algorithm as a high-quality candidate solution;

[0224] The genetic algorithm shares the excellent individuals in its population with the particle swarm algorithm and simulated annealing algorithm to increase the diversity of the population;

[0225] The simulated annealing algorithm shares the new solutions it accepts with the particle swarm algorithm and genetic algorithm to help escape from the local optimum;

[0226] The three algorithms update their respective parameters based on the shared information, such as particle speed, population individuals, temperature, etc.

[0227] Iterative optimization: The three updated algorithms enter the parallel optimization phase again and repeat steps BD until the maximum number of iterations or the accuracy requirement of the solution is met.

[0228] Termination output: When the termination condition is met, the load distribution solution with the lowest energy consumption is selected from the optimization results of the three algorithms as the final global optimal solution;

[0229] The optimal load distribution scheme is applied to the cold station control system. By adjusting the number of operating chillers, chilled water outlet temperature and other parameters, the cold station load is optimally configured to achieve energy-saving operation.

[0230] The embodiments of the present invention are given for the purpose of illustration and description, and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

Claims

1. A method for distributing cold station load based on an integrated optimization algorithm, characterized in that: The cooling station load distribution method based on the integrated optimization algorithm comprises the following steps: S1. Establish cold station model: By installing sensors on various subsystems and equipment in the cold station, the key parameters of the chiller's inlet and outlet water temperature, chilled water flow, and cooling capacity are collected in real time, and the data is transmitted to the database of the cold station control system for storage. On this basis, the characteristic curve of the chiller, the thermodynamic model of the cooling tower, and the water pump energy consumption model are integrated to establish a mathematical model that can accurately reflect the actual operating status of the cold station; S2. Constructing an optimization model for cooling station load distribution: Taking minimizing the total energy consumption of the cooling station as the objective function, combined with the operating condition constraints of the chiller and the balance conditions of the chilled water flow, an optimization model for the load distribution of the cooling station is constructed. The load distribution ratio of each cooling station is used as the decision variable, and the objective function value of each scheme is quantified through the energy consumption calculation formula. S3. Integrated optimization algorithm to solve the optimization model: The core idea of ​​the integrated optimization algorithm is to organically combine multiple heuristic algorithms and achieve collaborative optimization between algorithms through an information exchange mechanism. Specifically, the integrated optimization algorithm includes the following sub-steps: A: parameter initialization; B: Parallel optimization; C: Solution evaluation and information exchange; D: Algorithm update; E: Iterative solution; F: result output; The specific contents of parameter initialization are as follows: Initialize the common parameters of particle swarm optimization algorithm, genetic algorithm and simulated annealing algorithm uniformly, including initial solution set, fitness evaluation function and termination condition; The specific contents of parallel optimization are as follows: Particle swarm, genetic and simulated annealing algorithms are optimized in parallel in their respective search spaces, searching for the optimal load distribution solution from different angles. The particle swarm algorithm realizes the iteration of the solution by updating the position and velocity of particles, the genetic algorithm realizes the evolution of the population through selection, crossover and mutation operations, and the simulated annealing algorithm controls the probability of accepting a new solution through the temperature parameter. The specific contents of the evaluation and information exchange are as follows: Regularly conduct a unified fitness evaluation on the current solutions of the three algorithms, calculate the objective function value of each solution, identify the high-quality solutions of each algorithm based on the evaluation results, and exchange information through a shared memory mechanism. Among them, the particle swarm algorithm shares its global optimal solution with the genetic algorithm and joins the population as a new high-quality individual. The genetic algorithm shares the individual with the highest fitness in its population with the simulated annealing algorithm as the new current solution. The simulated annealing algorithm shares its accepted new solution with the particle swarm algorithm to help particles jump out of the local optimal area. The specific contents of the algorithm update are as follows: Based on the exchanged information, each algorithm adaptively updates its key parameters. The particle swarm algorithm adjusts the speed of particles according to the shared high-quality solutions, the genetic algorithm dynamically adjusts the crossover and mutation probabilities according to the external individuals introduced, and the simulated annealing algorithm adjusts the temperature decay rate according to the energy consumption level of the shared solutions. The specific contents of iterative solution are as follows: The three updated algorithms enter the parallel optimization phase again and repeat steps BD until the maximum number of iterations or the accuracy requirement of the solution is met. In each round of iteration, the algorithms continuously exchange information; The output results are as follows: When the termination condition is met, the load distribution scheme with the lowest total energy consumption is selected from the optimization results of the three algorithms as the final global optimal solution. The optimal solution is decoded into the load distribution ratio of each cooling station and transmitted to the next control module. S4. Application of optimal load distribution scheme: According to the optimal load distribution ratio, the operating parameter setting values ​​of each cold station and its internal chillers, the start and stop status of the chillers, the chilled water outlet temperature, and the chilled water flow rate are calculated. These control instructions are sent to the on-site controllers of each cold station through the data bus. The controller adjusts the operating status of the chillers according to the instructions to make the actual operating conditions close to the optimal solution. At the same time, the cold station control system also needs to periodically trigger the load distribution optimization program, obtain the operating condition changes of the cold station in real time, and dynamically update the optimization results to ensure that the cold station can achieve optimal energy consumption operation under different load demands.

2. The method for distributing cold station load based on integrated optimization algorithm according to claim 1, characterized in that: In step S2, according to the actual working conditions of the cold station, the range of the number of operating chillers and the upper and lower limits of the chilled water outlet temperature are set to ensure that the optimization result meets the feasibility requirements.

3. A method for distributing cold station load based on an integrated optimization algorithm according to any one of claims 1 to 2, characterized in that: The cooling station load distribution method based on the integrated optimization algorithm is applied to the technical field of cooling station load distribution.

Images