A smart optimization and energy-saving control algorithm and device for district cooling systems

Abstract

This invention provides an intelligent optimization and energy-saving control algorithm and device for a district cooling system, comprising: a control host, a chiller supply water temperature controller, a pump frequency controller, a chiller, a cooling tower, a pump, a water distributor, a water collector, and district cooling terminals. The control host includes: a CPU, a memory, an analog signal input interface, and an RS485 communication interface. The control host stores a novel intelligent optimization algorithm program based on the concept of national evolution processes. This algorithm possesses excellent global search capabilities in the early stages of iteration, solving the problem of existing optimization algorithms easily getting trapped in local optima. This invention also constructs an optimized energy-saving control model for the district cooling system and utilizes the intelligent optimization algorithm based on national evolution processes to achieve optimized energy-saving control of the district cooling system, maximizing the energy-saving effect of the district cooling system.

Description

Technical Field

[0001] This invention relates to the field of district cooling system optimization control technology, specifically to an intelligent optimization and energy-saving control algorithm and device for district cooling systems. Background Technology

[0002] With the rise of large building complexes such as commercial centers and transportation hubs, the application of district cooling and heating systems is becoming increasingly widespread. Optimizing and controlling district cooling systems to reduce energy consumption has significant economic and environmental benefits, making technological research in this area of ​​great importance.

[0003] One existing method involves using cleaned and processed historical system data to construct a neural network model, obtaining an overall system energy consumption model. Then, a cross-evolutionary algorithm is applied for optimization to determine the chilled water return temperature, chilled water supply temperature, chilled water flow rate, cooling water return temperature, cooling water supply temperature, and the number of chillers in operation, thus achieving energy-saving operation. However, because the neural network prediction model only relies on actual historical data to obtain the corresponding relationship model, it cannot maximize the reduction of system energy consumption, resulting in limited energy-saving effects. Another existing method involves establishing an energy consumption model for a district cooling system equipped with cold storage equipment, determining the return water temperature and supply water temperature range. However, the optimization algorithm used in this method is inefficient and cannot accurately identify the most energy-efficient operating conditions, resulting in poor optimization control. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent optimization and energy-saving control algorithm and device for district cooling systems, which solves the problem of poor energy consumption control in district cooling systems mentioned in the background.

[0006] (II) Technical Solution

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A smart optimization and energy-saving control device for a district cooling system includes: a control host, a chiller supply water temperature controller, a water pump frequency controller, a chiller, a cooling tower, a water pump, a water distributor, a water collector, and district cooling terminals.

[0009] The control host includes: a CPU, arithmetic logic unit, memory, an analog signal input interface, and an RS485 communication interface;

[0010] Temperature sensors are installed on the inlet and outlet sides of multiple chiller units and are connected to the control host through analog signal input interfaces;

[0011] The multiple pump frequency controllers are respectively connected to multiple pumps and connected to an RS485 communication interface;

[0012] The input and output terminals of the water pump frequency controller are connected to the RS485 communication interface and the water pump, respectively.

[0013] The water supply temperature controllers of the multiple chiller units are respectively connected to the multiple chiller units and connected to the RS485 communication interface;

[0014] The input and output terminals of the chiller unit's water supply temperature controller are connected to the RS485 communication interface and the chiller unit, respectively.

[0015] The control host stores a program for optimizing the control method of the district cooling system. The control host acquires data from the chiller unit's water supply temperature controller, water pump frequency controller, and temperature sensor.

[0016] The analog signal input interface supports 4~20 mA and 0~10V / 0~5V physical signals, and the RS485 communication interface supports Modbus RTU, Modbus ASCII, and PPI serial port protocols.

[0017] The present invention also provides an intelligent optimization and energy-saving control algorithm for a district cooling system, used to realize the function of an intelligent optimization and energy-saving control device for a district cooling system, characterized in that it includes: constructing an energy consumption model for the district cooling system;

[0018] An intelligent optimization algorithm based on national evolution processes is used to optimize energy efficiency in district cooling systems.

[0019] Preferably, the energy consumption model of the district cooling system includes: a water pump energy consumption model and a cooling unit energy consumption model;

[0020] The water pump energy consumption model includes: a chilled water pump model, a cooling water pump model, and a secondary pump model.

[0021] Preferably, the method for constructing the water pump energy consumption model includes:

[0022] The energy consumption model of the water pump is established by the following formula:

[0023]

[0024] In the formula: subscript Indicates a water pump; Indicates power, ; Indicates the fitting coefficients of the equipment model; Indicates chilled water mass flow rate Upper and lower limits, The correlation coefficients and constraints are provided by the equipment manufacturer.

[0025] Preferably, the energy consumption model of the cooling unit is constructed by means of:

[0026] The energy consumption model for chiller units is established using the following formula:

[0027]

[0028] In the formula: For unit energy consumption; For the unit's cooling load; The coefficient of performance (COP) of the generator unit;

[0029] The coefficient of performance (COP) of the generator unit is obtained by the following formula:

[0030] In the formula, r is the chiller unit load rate; This refers to the unit's condensing temperature. The evaporation temperature of the unit; , These are the fitting coefficients for the equipment model;

[0031] Unit condensing temperature We obtain it from the following formula:

[0032]

[0033] In the formula, This refers to the outlet temperature of the cooling water. For condenser load; This is the specific heat capacity of water; This refers to the cooling water flow rate; The overall heat transfer coefficient of the condenser is 250 W / (m²·K);

[0034] Unit evaporation temperature We obtain it from the following formula:

[0035]

[0036] In the formula, This refers to the outlet temperature of the chilled water. This refers to the chilled water flow rate; The overall heat transfer coefficient of the evaporator is 250 W / (m²·K).

[0037] Preferably, the energy consumption model for constructing the district cooling system is the sum of the water pump energy consumption model and the cooling unit energy consumption model;

[0038]

[0039] In the formula: This indicates the total energy consumption of the district cooling system. ; These represent the frequencies of each chilled water pump, each cooling water pump, and each secondary pump, respectively. ; This indicates the chilled water supply temperature for each chiller unit. ; Indicates the serial number of the chiller unit; Indicates the total number of chiller units; Indicates the serial number of the chilled water pump; Indicates the total number of chilled water pumps; Indicates the serial number of the cooling water pump; Indicates the total number of cooling water pumps; Indicates the secondary pump serial number; This indicates the total number of secondary pumps.

[0040] Preferably, the energy-saving optimization of the district cooling system using an intelligent optimization algorithm based on national evolution includes:

[0041] S1: Initialize the nation

[0042] Determine policy parameters for the initially established country, and set policy parameters ( Let D be the number of terms, and use equation (4) as the objective function to calculate the fitness values ​​of the initial policy parameters. The parameters of the initial policy can be expressed as:

[0043]

[0044] In the formula: This represents the value of the d-th dimension in the policy parameters;

[0045] S2: The State Determines Initial Policy Adjustments

[0046] The parameters are adjusted either significantly or slightly to generate a random number R uniformly distributed between (0,1). If R > 0.8, a significant adjustment is made; otherwise, a slight adjustment is made. The fitness value of the new parameters is then calculated after the policy parameters are adjusted.

[0047] Significant adjustments: New parameters will be randomly generated using Levy Flight, as follows:

[0048]

[0049]

[0050] In the formula: , ,generally ;

[0051] The new parameters are calculated as follows:

[0052]

[0053] like If the parameter exceeds the range, then:

[0054]

[0055] In the formula: For the new parameter in the d-th dimension, for The integer part, and These represent the upper and lower limits of the range for the d-th dimension data, respectively.

[0056] Minor adjustments: Based on the parameters corresponding to major powers, add requirements to meet... The random numbers are obtained from a normal distribution, and these are used as the new parameters, i.e.:

[0057]

[0058] In the formula: For the new parameter in the d-th dimension; Let d be the original parameter of dimension d; ;

[0059] like If the parameter exceeds the range, then:

[0060]

[0061] In the formula: For the new parameter in the d-th dimension, and These represent the upper and lower limits of the range for the d-th dimension data, respectively.

[0062] S3: Determine whether the adaptability of the initially adjusted policy parameters has improved.

[0063] If the parameter adaptability improves after the initial adjustment, proceed to step S2 to continue policy adjustments; otherwise, proceed to step S4.

[0064] S4: National Renewal Policy Parameters

[0065] The state uses the parameters that show the best adaptability in the initial policy adjustments as the new national policy.

[0066] S5: The country splits into smaller nations

[0067] A large country splits into n smaller countries. Each smaller country randomly retains some parameters from the large country, while the remaining parameters are slightly adjusted, using the same method as the slight adjustment in S2. The policy parameters established by the smaller countries can be expressed as:

[0068]

[0069] In the formula: This represents the policy parameter value of the i-th small country in the d-th dimension;

[0070] If the fitness of the newly generated parameters becomes worse than that of the large country, then the parameters of the small country will not be adjusted, and all parameters of the large country will be retained;

[0071] S6: Each small country appoints officials with specific duties.

[0072] Each small country has m officials, of whom 40% are ordinary courtiers, 40% are spies, and 20% are prime ministers;

[0073] The duties of each official are defined: ordinary courtiers offer suggestions and make policy adjustments. Spy agents are randomly dispatched to other small countries to learn from their courtiers or prime ministers about better policies. The prime minister listens extensively to the suggestions of his own courtiers and spies, learns from the best policies among them, and ultimately obtains the optimal policy parameters as the new policy parameters for the small countries.

[0074] Each official must adjust the policy parameters for each small country, that is, the initial parameters for each official are as shown in equation (12);

[0075] S7: Court officials propose policy adjustments.

[0076] The courtiers of each small kingdom began to offer suggestions, adjusting their respective kingdoms' policies accordingly. Each courtier generated a random number R uniformly distributed between (0,1). If R > 0.8, a large adjustment was made; otherwise, a small adjustment was made. The fitness value of the parameters was calculated after each adjustment.

[0077] For large-scale adjustments, new parameters are generated using Levy flight; for small-scale adjustments, new random numbers are generated using a normal distribution. The calculation steps are the same as in S2.

[0078] Each courtier makes c policy adjustments, and the parameters showing the best adaptability among each policy adjustment are recorded.

[0079] S8: The Spy's First Trip to Another Country for Study

[0080] Each spy from each small country randomly selects another small country and learns from the policy parameters proposed by the courtiers of that other country in a certain way, and calculates the fitness value of the learned parameters.

[0081] The Particle Swarm Optimization (PSO) algorithm is used for learning. The spy selects the best parameters among the courtiers of other countries as the learning object. That is, when the PSO algorithm particles are initialized, each particle has an 80% probability of using the learning object parameters as the particle's initial position and a 20% probability of using a random initial position. Then, PSO iterative calculation is performed, and the optimized best particle position is output as the new policy parameters that the spy can adopt after learning. The fitness of this parameter is compared with the fitness of the spy's original parameters. If it deteriorates, the spy does not adopt the suggestions of the courtiers of other countries, and the policy parameters are not changed.

[0082] The above process is repeated s times, meaning the spy randomly selects s ​​small countries for policy learning. Record the optimal parameter set for each small country after its initial learning.

[0083] S9: The Prime Minister's First Gathering of Opinions from Officials

[0084] Each prime minister of each small country selects the top 30% of the best policy parameters among the courtiers and spies of their respective small country as learning targets, and learns from them in a certain way to adjust the initial policy parameters of the small country.

[0085] The same PSO (Particle Swarm Optimization) algorithm as step S8 is used for learning. Each prime minister randomly selects a parameter from the top 30% of the best policy parameters among their own courtiers and spies as the learning target. PSO iterative calculations are performed, and the optimized best particle position is output as the prime minister's new policy parameter after learning. This process is repeated p times, and the best parameter obtained in each p-times of learning is used as the prime minister's new parameter. The initial set of best parameters learned by each small country's prime minister is recorded.

[0086] S10: The spy travels to another country for the second time to study.

[0087] Each spy from each small country randomly selects another small country and learns from the policy parameters of the prime minister of that country in a certain way. The fitness value of the learned parameters is calculated, and the PSO particle swarm algorithm is used for learning. The specific steps are the same as S8, except that the spy randomly selects the best parameter among the prime ministers of other countries as the learning object, and records the best parameter set of the second learning of each small country's spy.

[0088] S11: The Prime Minister gathers opinions from officials for the second time.

[0089] Each prime minister selects the top 30% of the best policy parameters from the courtiers and spies of their respective small country, and learns from them and adjusts the policy parameters in a certain way; the PSO particle swarm algorithm is used for learning, and the specific steps are the same as S9, and the best parameter set of each small country's prime minister is recorded for secondary learning;

[0090] S12: Obtain the best policy parameters among all small countries

[0091] Each small country updates its policy parameters: By comparing the fitness values ​​of the best parameter sets learned by the prime ministers of each small country during their initial and secondary learning processes, the best policy parameters among the prime ministers of each small country are obtained and used as the new policy parameters for each small country.

[0092] By comparing the new policy parameters of various small countries, the parameter with the best adaptability among all small countries can be expressed as:

[0093]

[0094] In the formula: This represents the value of the d-th dimension of the optimal policy parameter among all small countries;

[0095] S13: A small country with optimal policy parameters unifies all other countries.

[0096] Possessing optimal policy parameters A small country unifies with other countries and becomes a new great power, using its policy parameters as the initial policy parameters of the new country.

[0097] S14: Determine if the iteration termination condition has been met.

[0098] The algorithm terminates its iteration when the maximum number of iterations is reached or the specified convergence accuracy is reached. If the iteration termination condition is met, step S15 is executed; otherwise, step S2 is executed.

[0099] S15: Output optimal policy parameters

[0100] Through continuous national evolution, the optimal policy parameters are eventually determined. The largest country is output as the optimal value to complete the optimization calculation.

[0101] The optimal policy parameters are transmitted via RS485 communication interface to the chiller unit's water supply temperature controller and the water pump frequency controller to adjust the corresponding equipment parameters.

[0102] (III) Beneficial Effects

[0103] This invention provides an intelligent optimization and energy-saving control algorithm and device for a district cooling system, which has the following beneficial effects:

[0104] This invention proposes an intelligent optimization and energy-saving control algorithm and device for district cooling systems, comprising: constructing an energy consumption model for the district cooling system; employing an intelligent optimization algorithm based on national evolution processes to optimize the energy consumption of the district cooling system, achieving rapid convergence with a small number of iterations; utilizing mutual learning among small countries and between countries, enabling the algorithm to possess excellent global search capabilities in the early stages of iteration; and selecting superior learning targets to allow samples to learn from a better group, avoiding the problem of getting trapped in local optima; this invention has the ability to encompass multiple existing intelligent optimization algorithms, allowing the selection of appropriate algorithms based on the actual problem, enabling the algorithm to guide the system towards superior learning targets, thus solving the problem of poor energy consumption control in district cooling systems mentioned in the background. Attached Figure Description

[0105] Figure 1 A schematic diagram of the structure of an intelligent optimization and energy-saving control device for a district cooling system provided in an embodiment of the present invention;

[0106] Figure 2 A flowchart of an intelligent optimization and energy-saving control algorithm for a district cooling system provided in an embodiment of the present invention;

[0107] Figure 3 The flowchart illustrates a method for energy-saving optimization of a district cooling system using an intelligent optimization algorithm based on national evolution processes, provided in an embodiment of the present invention.

[0108] Figure 4 This is a comparison chart of energy consumption between an intelligent optimization and energy-saving control algorithm for a district cooling system provided in an embodiment of the present invention and other optimized algorithms.

[0109] In the diagram: 1. Control host; 2. CPU arithmetic unit; 3. Memory; 4. Analog signal input interface; 5. RS485 communication interface; 6. Chiller unit water supply temperature controller; 7. Water pump frequency controller; 8. Temperature sensor; 9. Chiller unit; 10. Cooling tower; 11. Water pump; 12. Water distributor; 13. Water collector; 14. District cooling terminal. Detailed Implementation

[0110] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0111] like Figure 1 As shown, this embodiment of the invention provides an intelligent optimization and energy-saving control device for a district cooling system, including: a control host 1, a chiller supply water temperature controller 6, a water pump frequency controller 7, a chiller 9, a cooling tower 10, a water pump 11, a water distributor 12, a water collector 13, and a district cooling terminal 14.

[0112] The control host 1 includes: a CPU arithmetic unit 2, a memory 3, an analog signal input interface 4, and an RS485 communication interface 5;

[0113] Temperature sensors 8 are installed on the inlet and outlet sides of multiple chiller units 9 and are connected to the control host 1 through analog signal input interface 4;

[0114] Multiple water pump frequency controllers 7 are respectively connected to multiple water pumps 11 and connected to RS485 communication interface 5;

[0115] The input and output terminals of the water pump frequency controller 7 are connected to the RS485 communication interface 5 and the water pump 11 respectively, so as to realize the frequency control of the water pump 11.

[0116] The multiple chiller unit water supply temperature controllers 6 are respectively connected to multiple chiller units 9 and connected to RS485 communication interface 5;

[0117] The input and output terminals of the chiller unit water supply temperature controller 6 are connected to the RS485 communication interface 5 and the chiller unit 9 respectively, so as to realize the control of the water supply temperature of the chiller unit 10.

[0118] The control host 1 stores a program for optimizing the control method of the district cooling system. The control host 1 acquires data from the chiller unit's water supply temperature controller 6, water pump frequency controller 7, and temperature sensor 8.

[0119] The analog signal input interface 4 supports 4~20 mA and 0~10V / 0~5V physical signals, and the RS485 communication interface supports Modbus RTU, Modbus ASCII, and PPI serial port protocols.

[0120] like Figure 2 As shown, this embodiment of the invention also provides an intelligent optimization and energy-saving control algorithm for a district cooling system, used to implement the function of an optimization and energy-saving control device for a district cooling system, including:

[0121] 201. Construct an energy consumption model for a district cooling system;

[0122] 202. An intelligent optimization algorithm based on national evolution process is used to optimize the energy efficiency of the district cooling system.

[0123] Preferably, the energy consumption model of the district cooling system includes: a water pump energy consumption model and a cooling unit energy consumption model;

[0124] The water pump energy consumption model includes: a chilled water pump model, a cooling water pump model, and a secondary pump model.

[0125] Preferably, the method for constructing the water pump energy consumption model includes:

[0126] The energy consumption model of the water pump is established by the following formula:

[0127]

[0128] In the formula: subscript Indicates a water pump; Indicates power, ; Indicates the fitting coefficients of the equipment model; Indicates chilled water mass flow rate Upper and lower limits, The correlation coefficients and constraints are provided by the equipment manufacturer.

[0129] Preferably, the energy consumption model of the cooling unit is constructed by means of:

[0130] The energy consumption model for chiller units is established using the following formula:

[0131]

[0132] In the formula: For unit energy consumption; For the unit's cooling load; The coefficient of performance (COP) of the generator unit;

[0133] The coefficient of performance (COP) of the generator unit is obtained by the following formula:

[0134] In the formula, r is the chiller unit load rate; This refers to the unit's condensing temperature. The evaporation temperature of the unit; , These are the fitting coefficients for the equipment model;

[0135] Unit condensing temperature We obtain it from the following formula:

[0136]

[0137] In the formula, This refers to the outlet temperature of the cooling water. For condenser load; This is the specific heat capacity of water; This refers to the cooling water flow rate; The overall heat transfer coefficient of the condenser is 250 W / (m²·K);

[0138] Unit evaporation temperature We obtain it from the following formula:

[0139]

[0140] In the formula, This refers to the outlet temperature of the chilled water. This refers to the chilled water flow rate; The overall heat transfer coefficient of the evaporator is 250 W / (m²·K).

[0141] Preferably, the energy consumption model for constructing the district cooling system is the sum of the water pump energy consumption model and the cooling unit energy consumption model;

[0142]

[0143] In the formula: This indicates the total energy consumption of the district cooling system. ; These represent the frequencies of each chilled water pump, each cooling water pump, and each secondary pump, respectively. ; This indicates the chilled water supply temperature for each chiller unit. ; Indicates the serial number of the chiller unit; Indicates the total number of chiller units; Indicates the serial number of the chilled water pump; Indicates the total number of chilled water pumps; Indicates the serial number of the cooling water pump; Indicates the total number of cooling water pumps; Indicates the secondary pump serial number; This indicates the total number of secondary pumps.

[0144] like Figure 3 As shown, the energy-saving optimization of the district cooling system using an intelligent optimization algorithm based on national evolution includes:

[0145] S1: Initialize the nation

[0146] Determine policy parameters for the initially established country, and set policy parameters ( Let D be the number of terms, and use equation (4) as the objective function to calculate the fitness values ​​of the initial policy parameters. The parameters of the initial policy can be expressed as:

[0147]

[0148] In the formula: This represents the value of the d-th dimension in the policy parameters;

[0149] S2: The State Determines Initial Policy Adjustments

[0150] The parameters are adjusted either significantly or slightly to generate a random number R uniformly distributed between (0,1). If R > 0.8, a significant adjustment is made; otherwise, a slight adjustment is made. The fitness value of the new parameters is then calculated after the policy parameters are adjusted.

[0151] Significant adjustments: New parameters will be randomly generated using Levy Flight, as follows:

[0152]

[0153]

[0154] In the formula: , ,generally ;

[0155] The new parameters are calculated as follows:

[0156]

[0157] like If the parameter exceeds the range, then:

[0158]

[0159] In the formula: For the new parameter in the d-th dimension, for The integer part, and These represent the upper and lower limits of the range for the d-th dimension data, respectively.

[0160] Minor adjustments: Based on the parameters corresponding to major powers, add requirements to meet... The random numbers are obtained from a normal distribution, and these are used as the new parameters, i.e.:

[0161]

[0162] In the formula: For the new parameter in the d-th dimension; Let d be the original parameter of dimension d; ;

[0163] like If the parameter exceeds the range, then:

[0164]

[0165] In the formula: For the new parameter in the d-th dimension, and These represent the upper and lower limits of the range for the d-th dimension data, respectively.

[0166] S3: Determine whether the adaptability of the initially adjusted policy parameters has improved.

[0167] If the parameter adaptability improves after the initial adjustment, proceed to step S2 to continue policy adjustments; otherwise, proceed to step S4.

[0168] S4: National Renewal Policy Parameters

[0169] The state uses the parameters that show the best adaptability in the initial policy adjustments as the new national policy.

[0170] S5: The country splits into smaller nations

[0171] A large country splits into n smaller countries. Each smaller country randomly retains some parameters from the large country, while the remaining parameters are slightly adjusted, using the same method as the slight adjustment in S2. The policy parameters established by the smaller countries can be expressed as:

[0172]

[0173] In the formula: This represents the policy parameter value of the i-th small country in the d-th dimension;

[0174] If the fitness of the newly generated parameters becomes worse than that of the large country, then the parameters of the small country will not be adjusted, and all parameters of the large country will be retained;

[0175] S6: Each small country appoints officials with specific duties.

[0176] Each small country has m officials, of whom 40% are ordinary courtiers, 40% are spies, and 20% are prime ministers;

[0177] The duties of each official are defined: ordinary courtiers offer suggestions and make policy adjustments. Spy agents are randomly dispatched to other small countries to learn from their courtiers or prime ministers about better policies. The prime minister listens extensively to the suggestions of his own courtiers and spies, learns from the best policies among them, and ultimately obtains the optimal policy parameters as the new policy parameters for the small countries.

[0178] Each official must adjust the policy parameters for each small country, that is, the initial parameters for each official are as shown in equation (12);

[0179] S7: Court officials propose policy adjustments.

[0180] The courtiers of each small kingdom began to offer suggestions, adjusting their respective kingdoms' policies accordingly. Each courtier generated a random number R uniformly distributed between (0,1). If R > 0.8, a large adjustment was made; otherwise, a small adjustment was made. The fitness value of the parameters was calculated after each adjustment.

[0181] For large-scale adjustments, new parameters are generated using Levy flight; for small-scale adjustments, new random numbers are generated using a normal distribution. The calculation steps are the same as in S2.

[0182] Each courtier makes c policy adjustments, and the parameters showing the best adaptability among each policy adjustment are recorded.

[0183] S8: The Spy's First Trip to Another Country for Study

[0184] Each spy from each small country randomly selects another small country and learns from the policy parameters proposed by the courtiers of that other country in a certain way, and calculates the fitness value of the learned parameters.

[0185] The Particle Swarm Optimization (PSO) algorithm is used for learning. The spy selects the best parameters among the courtiers of other countries as the learning object. That is, when the PSO algorithm particles are initialized, each particle has an 80% probability of using the learning object parameters as the particle's initial position and a 20% probability of using a random initial position. Then, PSO iterative calculation is performed, and the optimized best particle position is output as the new policy parameters that the spy can adopt after learning. The fitness of this parameter is compared with the fitness of the spy's original parameters. If it deteriorates, the spy does not adopt the suggestions of the courtiers of other countries, and the policy parameters are not changed.

[0186] The above process is repeated s times, meaning the spy randomly selects s ​​small countries for policy learning. Record the optimal parameter set for each small country after its initial learning.

[0187] S9: The Prime Minister's First Gathering of Opinions from Officials

[0188] Each prime minister of each small country selects the top 30% of the best policy parameters among the courtiers and spies of their respective small country as learning targets, and learns from them in a certain way to adjust the initial policy parameters of the small country.

[0189] The same PSO (Particle Swarm Optimization) algorithm as step S8 is used for learning. Each prime minister randomly selects a parameter from the top 30% of the best policy parameters among their own courtiers and spies as the learning target. PSO iterative calculations are performed, and the optimized best particle position is output as the prime minister's new policy parameter after learning. This process is repeated p times, and the best parameter obtained in each p-times of learning is used as the prime minister's new parameter. The initial set of best parameters learned by each small country's prime minister is recorded.

[0190] S10: The spy travels to another country for the second time to study.

[0191] Each spy from each small country randomly selects another small country and learns from the policy parameters of the prime minister of that country in a certain way. The fitness value of the learned parameters is calculated, and the PSO particle swarm algorithm is used for learning. The specific steps are the same as S8, except that the spy randomly selects the best parameter among the prime ministers of other countries as the learning object, and records the best parameter set of the second learning of each small country's spy.

[0192] S11: The Prime Minister gathers opinions from officials for the second time.

[0193] Each prime minister selects the top 30% of the best policy parameters from the courtiers and spies of their respective small country, and learns from them and adjusts the policy parameters in a certain way; the PSO particle swarm algorithm is used for learning, and the specific steps are the same as S9, and the best parameter set of each small country's prime minister is recorded for secondary learning;

[0194] S12: Obtain the best policy parameters among all small countries

[0195] Each small country updates its policy parameters: By comparing the fitness values ​​of the best parameter sets learned by the prime ministers of each small country during their initial and secondary learning processes, the best policy parameters among the prime ministers of each small country are obtained and used as the new policy parameters for each small country.

[0196] By comparing the new policy parameters of various small countries, the parameter with the best adaptability among all small countries can be expressed as:

[0197]

[0198] In the formula: This represents the value of the d-th dimension of the optimal policy parameter among all small countries;

[0199] S13: A small country with optimal policy parameters unifies all other countries.

[0200] Possessing optimal policy parameters A small country unifies with other countries and becomes a new great power, using its policy parameters as the initial policy parameters of the new country.

[0201] S14: Determine if the iteration termination condition has been met.

[0202] The algorithm terminates its iteration when the maximum number of iterations is reached or the specified convergence accuracy is reached. If the iteration termination condition is met, step S15 is executed; otherwise, step S2 is executed.

[0203] S15: Output optimal policy parameters

[0204] Through continuous national evolution, the optimal policy parameters are eventually determined. The largest country is output as the optimal value to complete the optimization calculation.

[0205] The optimal parameters are transmitted via RS485 communication interface 5 to the chiller unit water supply temperature controller 6 and the water pump frequency controller 7 to adjust the corresponding equipment parameters.

[0206] To verify the control algorithm provided in this embodiment of the invention, taking an example consisting of 3 chiller units, 4 chilled water pumps, 8 secondary pumps, 4 cooling water pumps, and 4 cooling towers, the parameters of each device are shown in the table below:

[0207] Table 1 Chiller Unit Parameters

[0208]

[0209] Table 2 Pump Parameters

[0210]

[0211] Water pump fitting formula: .

[0212] Under conditions of an outdoor temperature of 34.5℃, an outdoor relative humidity of 72%, and an estimated cooling load demand of 4000kW, the system energy consumption using the traditional fixed-parameter control method is 1100.72kW. After optimization using intelligent algorithms such as Particle Swarm Optimization (PSO), Bluebottle Bowerbird Optimization (SBO), Sparrow Search Algorithm (SSA), and Country Evolution Algorithm (CEA), the system energy consumption is reduced to 801.36kW, 789.14kW, 830.98kW, and 777.11kW, respectively. Figure 4 As shown in the figure, the optimization results show that the energy-saving effect of the district cooling system optimization control method based on the National Evolution Process (CEA) provided by the embodiment of the present invention is the best, reducing system energy consumption by 323.61kW compared with the traditional control method, which is better than the SBO algorithm.

[0213] In summary, the National Evolution Process (CEA) provided by the embodiments of the present invention will have the best optimization effect in the intelligent optimization control of district cooling systems.

[0214] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent optimization energy-saving control algorithm for district cooling systems, characterized in that, The application relates to a method for energy-saving optimization of a district cooling system. The method comprises the following steps: S1: initializing a country S2: determining a preliminary policy adjustment of the country S3: judging whether the fitness of the policy parameter after the preliminary adjustment is improved Determine policy parameters for the initially established country, and set policy parameters ( The number of ) is D, and with (4) as the objective function, wherein, Tcondis the condensing temperature of the unit , Toutis the outlet temperature of the cooling water; Qcondis the condensing load; Cwis the specific heat capacity of water; Fwis the flow rate of the cooling water; Ucondis the overall heat transfer coefficient of the condenser, which is 250 W / (m2·K); calculate the fitness value of the initial policy parameters; the parameters of the initial policy can be represented as: （7） In the formulae: denotes the value of the d-th dimension in the policy parameters; S4: updating the policy parameter of the country S5: splitting the country into small countries S6: setting up officials in each small country （8） （9） In the formulae: , , generally ; S7: making suggestions for policy adjustment by the officials （10） If If the parameter range is exceeded, then there is: （11） wherein: is the d-th dimension new parameter, is is the integer part of and are the upper and lower range limits of the d-th dimension data, respectively. Minor adjustments: Based on the parameters corresponding to major powers, add requirements to meet... The random numbers are obtained from a normal distribution, and these are used as the new parameters, i.e.: （12） In the formulae: is the d-th new parameter; is the d-th old parameter; ; If If the parameter range is exceeded, then there is: （13） wherein: is the dth new parameter, and are the upper and lower range limits, respectively, for the dth dimension of data. S8: learning from other countries by the officials ​ ​ ​ ​ ​ （14） wherein: represents the value of the policy parameter in dimension d for the ith small country; ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ The PSO particle swarm algorithm is used for learning, and the best parameters of the officials of other countries are selected as learning objects, that is, when the PSO algorithm particles are initialized, each particle has a probability of 80% to take the learning object parameters as the initial position of the particle, and a probability of 20% to take a random initial position, and then the PSO iterative calculation is performed, and the optimized best particle position is output as the new policy parameter that can be adopted by the explorer after learning, and the fitness of the parameter is compared with the fitness of the original parameter of the explorer, if the fitness is worse, the explorer does not adopt the suggestion of the officials of other countries, and the policy parameter is not changed; The above process is repeated s times, that is, the explorer randomly selects s small countries for policy learning; and the best parameter set of each small country explorer after initial learning is recorded; S9: The prime minister initially collects the opinions of the officials Each prime minister of each small country selects the top 30% of the best policy parameters of the officials and explorers in the small country as learning objects, and learns from them in a certain way to adjust the initial policy parameters of the small country; The same PSO particle swarm algorithm as in step S8 is used for learning, each prime minister randomly selects a parameter in the top 30% of the best policy parameters of the officials and explorers in the country as a learning object, performs PSO iterative calculation, and outputs the optimized best particle position as the new policy parameter of the prime minister after learning, and this process is repeated p times, the best parameters obtained in the p learning processes are taken as the new parameters of the prime minister; and the best parameter set of each small country prime minister after initial learning is recorded; S10: The explorer visits other countries for the second time Each explorer of each small country randomly selects another small country, learns from the policy parameters of the prime minister of the other country in a certain way, calculates the fitness value of the learned parameters, uses the PSO particle swarm algorithm for learning, and the specific steps are the same as in S8, except that the explorer randomly selects the best parameter of the prime minister of the other country as the learning object, and the best parameter set of each small country explorer after the second learning is recorded; S11: The prime minister collects the opinions of the officials for the second time Each prime minister selects the top 30% of the best policy parameter set of the officials and explorers in the small country, learns from them in a certain way, and adjusts the policy parameters; the PSO particle swarm algorithm is used for learning, and the specific steps are the same as in S9, and the best parameter set of each small country prime minister after the second learning is recorded; S12: Obtain the best policy parameters in all small countries Each small country updates the policy parameters: by comparing the fitness values of the best parameter sets of the prime ministers of each small country after initial and secondary learning, the best policy parameters in the prime ministers of each small country are obtained, and the best policy parameters in the prime ministers of each small country are taken as the new policy parameters of each small country; The new policy parameters of each small country are compared, and the best parameters in all small countries are obtained, which can be represented as: （15） where: represents the value of the d-th dimension of the best policy parameter among all small countries; S13: The small country with the best policy parameters unifies all countries has the best policy parameters unifies the countries with the best policy parameters into a new country, which becomes a new large country, and uses the policy parameters of the new country as the initial policy parameters of the new large country; S14: Determine whether the iteration termination condition is reached The iteration termination condition of the algorithm is that the maximum number of iterations is reached or the specified convergence precision is reached, if the iteration termination condition is reached, step S15 is performed, otherwise, step S2 is performed; S15: Output the best policy parameters After the continuous evolution of the country, eventually have the best policy parameters The great power as the optimal value output, complete optimization calculation.

2. The intelligent optimization and energy-saving control algorithm for a district cooling system according to claim 1, characterized in that, The regional cooling system energy consumption model comprises a water pump energy consumption model and a cooling unit energy consumption model. The water pump energy consumption model comprises a chilled water pump model, a cooling water pump model, and a secondary pump model.

3. The intelligent optimization and energy-saving control algorithm for a district cooling system according to claim 2, characterized in that, The water pump energy consumption model is constructed by the following method: The energy consumption model of the water pump is established by the following formula: （1） In the formula: subscript Indicates a water pump; Indicates power, ; Indicates the fitting coefficients of the equipment model; Indicates chilled water mass flow rate Upper and lower limits, The correlation coefficients and constraints are provided by the equipment manufacturer.

4. The intelligent optimization and energy-saving control algorithm for a district cooling system according to claim 3, characterized in that, The energy consumption model of the cooling unit, the construction method comprises: The energy consumption model of the water chiller is established by the following formula: （2） In the formula: is the energy consumption of the unit; is the cooling load of the unit; is the coefficient of performance of the unit; The coefficient of performance COP of the unit is obtained by the following formula: (3) In the formula, r is a load rate of the water chiller; is a condensing temperature of the unit; is an evaporating temperature of the unit; 、 is a fitting coefficient of the equipment model; Unit condensing temperature By the following formula: （4） wherein Tout is the cooling water outlet temperature; Qcond is the condenser load; Cp is the specific heat capacity of water; F is the cooling water flow rate; U is the overall heat transfer coefficient of the condenser, which is 250 W / (m2·K); Unit evaporation temperature By the following formula: （5） wherein Tout is the chilled water outlet temperature; G is the chilled water flow rate; U is the total heat transfer coefficient of the evaporator, and is 250 W / (m2·K).

5. The intelligent optimization and energy-saving control algorithm for a district cooling system according to claim 4, characterized in that, The energy consumption model of the regional cooling system is the sum of the energy consumption model of the water pump and the energy consumption model of the cooling unit; （6） In the formula: represents the total energy consumption of the district cooling system, ; represents the frequency of each chilled water pump, each cooling water pump, and each secondary pump, respectively, ; represents the chilled water supply temperature of each chiller unit, ; represents the chiller unit serial number; represents the total number of chiller units; represents the chilled water pump serial number; represents the total number of chilled water pumps; represents the cooling water pump serial number; represents the total number of cooling water pumps; represents the secondary pump serial number; represents the total number of secondary pumps.

Images