An intelligent optimization control method for a power station house cold source system of a factory

Abstract

The application provides an intelligent optimization control method for a power station house cold source system, comprising the following steps: firstly, the mechanism of the cold source system and the actual operation big data of the factory are combined to establish an energy-saving optimization control model for the power station house cold source system; secondly, the operation mechanism and the energy consumption expression of the cold machine are analyzed in depth, the properties of the problems to be optimized are determined, and the equality or inequality constraint conditions are established according to the inherent properties of the cold machine equipment and the industrial site environment; thirdly, a particle swarm optimization algorithm with shared knowledge files is designed, a learning optimization model is created to guide the direction of group intelligence search, the optimal branch chilled water inlet / outlet water temperature after decoding is calculated and solved, and finally, the control part is provided, and the COP (Coefficient Of Performance) index is significantly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent optimization control, and in particular to an intelligent optimization control method for a cold source system of a power station in a factory. Background Art

[0002] Currently, the cooling system is one of the most important components of central air conditioning systems, serving as the source of the air conditioning system. The main unit's energy consumption accounts for approximately 60% to 70% of the total system's energy consumption, while the refrigeration and cooling pumps contribute approximately 20% to 30%. Proper control and optimized operation of these systems can achieve significant energy savings. Central air conditioning cooling system equipment includes chillers, chilled water pumps, cooling water pumps, and cooling towers. The parameters between these devices are coupled and exhibit complex relationships, such as nonlinear and time-varying properties. Therefore, establishing an appropriate energy consumption model for central air conditioning cooling system equipment, analyzing their mutual constraints and influences, and constructing an efficient intelligent optimization model are key to achieving energy-efficient operation of central air conditioning cooling systems. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent optimization control method for a factory power station cooling system, which can effectively reduce enterprise costs and achieve efficient and energy-saving operation of the cooling system.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] An intelligent optimization control method for a cooling system of a power station in a factory comprises the following steps:

[0006] Step S1: Build a system optimization model, combining the industrial system mechanism with actual operation big data to establish an energy-saving optimization control model for the factory refrigeration equipment system;

[0007] Step S2: Determine the decision variables, analyze the energy consumption expression of the energy-saving optimization control model, and determine the nature of the problem to be optimized;

[0008] Step S3: setting constraints and analyzing the constraints based on the inherent properties of the refrigeration equipment, on-site production conditions, and requirements for ensuring safe and stable operation of the refrigeration equipment;

[0009] Step S4: Use a swarm intelligence optimization algorithm based on a shared knowledge archive to solve and output the optimal control parameters to the cooling system actuator.

[0010] Preferably, the energy-saving optimization control model in step S1 establishes a mathematical model of the cold source system based on actual operating data of the refrigeration equipment, and the refrigeration equipment includes a refrigeration unit, a chilled water pump, and a cooling water pump.

[0011] Preferably, the energy-saving optimization control model is as follows:

[0012] (2)

[0013] in, is the actual input power / power consumption of a single chiller, is the number of cooling units turned on, is the actual cooling capacity of a single chiller, is the efficiency coefficient of a single chiller, is the density of water, is the chilled water flow rate of a single chiller, is the specific heat capacity of water, is the chilled water inlet temperature of the main pipe, is the chilled water outlet temperature of a single chiller, is the outlet temperature of chilled water from the main pipe, is the total terminal load demand, is the chilled water flow of the chiller.

[0014] Preferably, the properties of the problem to be optimized in step S2 include analyzing the actual operating data of the refrigeration equipment and the cold source system mechanism, and selecting the chilled water outlet temperature, cooling water inlet temperature, chilled water pump flow, cooling water pump flow, etc. of the refrigeration unit as optimization control parameters.

[0015] Preferably, the constraint conditions described in step S3 include obtaining data on the state of the refrigeration machine being on or off, the total load demand value, the time when each refrigeration machine last changed its state, the activation state of the refrigeration machine, and the accumulated startup time, to determine the selection and screening conditions of the refrigeration machine control scheme.

[0016] Preferably, the constraint conditions are obtained as follows:

[0017] (3)

[0018] in, is the rated cooling capacity of a single chiller, and are the lower and upper limits of the cooling machine load rate, and They are the lower and upper limits of chilled water supply temperature for the chiller respectively.

[0019] Preferably, the constraint conditions include dynamic adjustment constraints, and the dynamic adjustment constraints include: when a new control plan is generated based on the current refrigeration equipment startup plan, the number of refrigeration equipment started and stopped does not change by more than 2 units, and the total cooling capacity change of the system does not exceed the rated value of the single maximum rated cooling capacity of the centrifuge refrigeration machine.

[0020] Preferably, the particle swarm optimization algorithm based on the shared knowledge archive in step S4 implements intelligent optimization control through the following steps:

[0021] Step S41: Divide the population into subpopulations based on historical experience information, and set the subpopulation adaptation function to dynamically allocate computing resources.

[0022] Step S42: Establish a global archive set to store co-evolution information, and realize self-organization update of elite particles through horizontal individual communication among sub-populations;

[0023] Step S43: Build a learning optimization model and use the historical optimal information of the archive set to guide the crowd-intelligence search;

[0024] Step S44: The population is gradually converged through iterative evolution, and the optimal feasible solution that meets the constraints, i.e., the control scheme, is finally output to achieve global intelligent optimization control.

[0025] Preferably, the specific steps are as follows:

[0026] First, an initial population (100 individuals) is randomly generated in the decision space and divided into subpopulations (10 subpopulations, each with 10 individuals). By comparing the energy consumption values ​​of individuals, some (20 individuals) of excellent individuals are selected and put into the archive set, and a subpopulation individual update method is designed.

[0027] (4)

[0028] in, is a new individual in the subpopulation, is the original individual of the subpopulation, is the particle velocity, is the optimal individual in the archive set, and Randomly select individuals from the adjacent populations of the sub-population (judgment by sub-population number), , and is the coefficient and . Create a learning optimization model based on historical knowledge archives and output , and Construct new individuals to guide the direction of collective intelligence search;

[0029] If the new individual is not feasible, then the new individual is updated repeatedly through formula (4); if the same new individual is not feasible multiple times, then a random value is assigned in the feasible domain;

[0030] Each subpopulation generates an equal number of viable new individuals, and an equal number of next-generation individuals are selected through fitness. The entire population then selects outstanding individuals for archiving, and through algorithmic iterations, the overall population evolution is achieved. Upon reaching the corresponding algorithm termination condition, the population provides an optimized solution to the problem. This solution is decoded to obtain the optimal branch chilled water inlet and outlet temperatures in the cooling system, which are ultimately output to the control system, significantly improving the COP.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. Significantly improve system energy efficiency and reduce energy consumption costs

[0033] This method combines a mechanism model with big data analysis to establish an optimized control model, precisely matching actual load demands with equipment operating parameters. It dynamically optimizes key variables such as chilled water outlet temperature and cooling water inlet temperature, ensuring the system consistently operates within the optimal Cost-Effectiveness Ratio (COP) range. Experimental data demonstrates that this method can reduce the overall energy consumption of the cooling system by 10% to 25%, significantly reducing corporate electricity costs. It is particularly suitable for high-energy-consuming manufacturing industries.

[0034] 2. Ensure equipment safety and stability

[0035] The present invention avoids drastic fluctuations in control instructions through rigid specifications of constraints (such as the time interval between refrigeration unit start-up and shutdown, the limit on the number of units changed, and the cooling capacity fluctuation threshold), reduces frequent start-up and shutdown or overload operation of equipment, and extends the service life of core equipment such as refrigeration units and water pumps.

[0036] Combined with cold machine status monitoring (accumulated startup time, activation status, etc.), preventive maintenance can be achieved to reduce the risk of unexpected downtime.

[0037] 3. Dynamically adapt to complex working conditions and strong robustness

[0038] The present invention adopts a strategy that combines static planning with real-time data feedback. Each round of optimization generates a solution based on the current operating condition data. This can not only quickly respond to load changes (such as production schedule adjustments and ambient temperature fluctuations), but also avoid the high computational complexity of traditional dynamic optimization algorithms.

[0039] Through subpopulation division and global knowledge archiving mechanisms, the algorithm can adapt to multi-objective conflicts (such as energy efficiency priority vs. equipment life balance) and meet the needs of different factories.

[0040] 4. The optimization algorithm converges efficiently and is suitable for industrial real-time control

[0041] The present invention guides subpopulation search through historical experience, reduces invalid iterations, and improves convergence speed by 30% to 50% compared to traditional optimization methods (such as genetic algorithms), meeting the real-time requirements of factory control systems (usually responding in seconds).

[0042] 5. Reduce manual intervention and improve automation

[0043] The present invention replaces the traditional parameter adjustment that relies on manual experience, realizes fully automatic closed-loop control, and reduces the technical threshold for operation and maintenance personnel.

[0044] Through the accumulation of long-term operating data, the system can autonomously learn the factory's energy consumption patterns and further optimize the control strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of an intelligent optimization control method for a factory power station cooling system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0047] like Figure 1 As shown in FIG, the intelligent optimization control method for the cooling source system of a power station in a factory provided in this embodiment has the following steps:

[0048] (1) A method that combines industrial system mechanisms with actual operation big data to establish an optimization control model for factory cooling systems.

[0049] Based on the actual operating data of cold source equipment such as refrigeration units, chilled water pumps, and cooling water pumps, a mathematical model of the cold source system is established. The chilled water outlet temperature, cooling water inlet temperature, chilled water pump flow rate, and cooling water pump flow rate of the refrigeration unit are selected as the main optimization control parameters. The solution range is established for the decision variables, that is, the upper and lower limits of the control parameters are determined to ensure stable and efficient operation of the system. The ultimate goal of HVAC system optimization control is to reduce the energy consumption of the cold source system, that is, to improve the system's energy efficiency ratio (COP, Coefficient Of Performance), which can be expressed as,

[0050] (1)

[0051] in, is the number of chillers in the air conditioning cold source system, For the The power of the chiller, is the number of chilled water pumps, For the Energy consumption of a chilled water pump, is the number of cooling water pumps, For the Energy consumption of cooling water pumps, For heat, is the total energy consumption.

[0052] (2) Determine the nature of the problem to be optimized

[0053] Based on the chiller energy consumption expression from the data modeling section and considering the actual requirements for optimized operation of the project, this problem can be determined to be a deterministic optimization problem. Furthermore, since the feasible solution is not a function of time—that is, once the field data is returned at a certain moment, the feasible solution does not change over time—this optimization problem is therefore considered a static planning problem.

[0054] (3) Constraint analysis

[0055] By reading the data sent back by the control end at the time of the request in the database, the optimization module obtains the current status of all chillers on or off, the total load demand value, the time when each chiller last changed its status, the chiller activation status, the cumulative startup time and other data as the selection and screening conditions for feasible control schemes for the chiller.

[0056] Based on the unit status and the requirements for ensuring safe and stable operation of the chiller, the model establishes the upper limit of the total terminal load and the chiller temperature control range. Furthermore, it is necessary to ensure that the chiller startup plan does not undergo drastic changes during the plan adjustment and optimization process, otherwise it will increase the possibility of instability or even damage to the chiller control system. Therefore, all feasible solutions based on the original startup plan must meet the special constraints of no more than two units change and the change in total customized cooling capacity does not exceed the rated cooling capacity of a single large centrifuge.

[0057] (4) Design and operation of optimization algorithms

[0058] After determining the nature of the problem, target quantity, constraints and other related information, it is necessary to identify each chiller with four state labels: openable, not openable, closeable, and not closeable, based on the mechanism characteristics of the factory power station cooling system and relevant conditions such as activation state and state change time. It is also necessary to establish the encoding and reverse decoding methods of decision variables from actual physical space to operable mathematical space.

[0059] A particle swarm optimization algorithm based on a shared knowledge archive is designed. Subpopulations are divided according to historical experience, and computing resources are dynamically allocated based on subpopulation fitness. A global knowledge archive is introduced to store co-evolutionary information. Transverse communication between subpopulations enables efficient updates of elite particle heuristic self-organization. A learning optimization model is created to guide swarm intelligence search. A cyclical and iterative population evolution mechanism is used to gradually push each decision vector toward the global optimal direction, achieving overall evolution of the swarm. Upon reaching the corresponding algorithm termination condition, the swarm provides an optimized solution to the problem. Through decoding, the optimal branch chilled water inlet and outlet temperatures in the cooling system are determined and ultimately transmitted to the control system, thereby achieving optimized control of the cooling system in the power station building and significantly improving the cost-effectiveness (COP).

[0060] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0061] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0062] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. An intelligent optimization control method for a factory power station cooling system, characterized in that: The steps include: Step S1: Build a system optimization model, combining the industrial system mechanism with actual operation big data to establish an energy-saving optimization control model for the factory refrigeration equipment system; In step S1, the energy-saving optimization control model establishes a mathematical model of the cold source system based on the actual operating data of the refrigeration equipment, wherein the refrigeration equipment includes a refrigeration unit, a chilled water pump, and a cooling water pump; The energy-saving optimization control model is as follows: (2) in, is the actual input power / power consumption of a single chiller, is the number of cooling units turned on, is the actual cooling capacity of a single chiller, is the efficiency coefficient of a single chiller, is the density of water, is the chilled water flow rate of a single chiller, is the specific heat capacity of water, is the chilled water inlet temperature of the main pipe, is the chilled water outlet temperature of a single chiller, is the outlet temperature of chilled water from the main pipe, is the total terminal load demand, is the chilled water flow of the chiller; Step S2: Determine the decision variables, analyze the energy consumption expression of the energy-saving optimization control model, and determine the nature of the problem to be optimized; Step S3: setting constraints and analyzing the constraints based on the inherent properties of the refrigeration equipment, on-site production conditions, and requirements for ensuring safe and stable operation of the refrigeration equipment; The constraint conditions in step S3 include obtaining data on the on / off status of the chiller, the total load demand value, the time when each chiller last changed its status, the chiller activation status, and the accumulated startup time, to determine the selection and screening conditions of the chiller control scheme; Step S4: using a swarm intelligence optimization algorithm based on a shared knowledge archive to solve and output the optimal control parameters to the cooling system actuator; The particle swarm optimization algorithm based on the shared knowledge archive in step S4 implements intelligent optimization control through the following steps: Step S41: Divide the population into subpopulations based on historical experience information, and set the subpopulation adaptation function to dynamically allocate computing resources. Step S42: establishing a global archive set to store the co-evolution information, and realizing the self-organization update of elite particles through horizontal individual communication among sub-populations; Step S43: Build a learning optimization model and use the historical optimal information of the archive set to guide the crowd-intelligence search; Step S44: The population is gradually converged through iterative evolution, and the optimal feasible solution that meets the constraints, i.e., the control scheme, is finally output to achieve global intelligent optimization control.

2. The intelligent optimization control method for the cooling source system of a power station in a factory according to claim 1 is characterized in that: The properties of the problem to be optimized in step S2 include analyzing the actual operating data of the refrigeration equipment and the cold source system mechanism, and selecting the chilled water outlet temperature, cooling water inlet temperature, chilled water pump flow rate, and cooling water pump flow rate of the refrigeration unit as optimization control parameters.

3. The intelligent optimization control method for the cooling source system of a power station in a factory according to claim 1 is characterized in that: The constraints are obtained as follows: (3) in, is the rated cooling capacity of a single chiller, and are the lower and upper limits of the cooling machine load rate, and They are the lower and upper limits of chilled water supply temperature for the chiller respectively.

4. The intelligent optimization control method for the cooling source system of a power station in a factory according to claim 1 is characterized in that: The constraint conditions include dynamic adjustment constraints, which include: when generating a new control plan based on the current refrigeration equipment startup plan, the number of refrigeration equipment started and stopped shall not change by more than 2 units, and the total cooling capacity change of the system shall not exceed the rated value of the single maximum rated cooling capacity of the centrifuge refrigeration machine.

5. The intelligent optimization control method for the cooling source system of a power station in a factory according to claim 1 is characterized in that: The specific steps are as follows: First, the initial population is randomly generated in the decision space and divided into subpopulations. By comparing the energy consumption values ​​of individuals, excellent individuals are selected to enter the archive set, and the subpopulation individual update method is designed. (4) in, is a new individual in the subpopulation, is the original individual of the subpopulation, is the particle velocity, is the optimal individual in the archive set, and Randomly select individuals from the adjacent populations of the subpopulation, , and is the coefficient and Create a learning optimization model based on historical knowledge archives and output , and Construct new individuals to guide the direction of collective intelligence search; If the new individual is not feasible, then the new individual is updated repeatedly through formula (4); if the same new individual is not feasible multiple times, then a random value is assigned in the feasible domain; Each subpopulation will produce an equal number of feasible new individuals and screen out an equal number of next-generation individuals through fitness. The entire population will then select outstanding individuals to enter the archive set, and through the cyclic iteration of the algorithm, the overall evolution of the population will be achieved. After reaching the corresponding algorithm termination condition, the population will provide an optimized solution to the problem. After decoding, the optimized branch chilled water inlet / outlet temperature in the cold source system is obtained, and finally output to the control part, significantly improving the COP index.

Images