Multi-target operation optimization method for combined cooling heating and power comprehensive energy system

Abstract

The invention belongs to the technical field of energy system optimization, and provides a combined cooling heating and power comprehensive energy system multi-target operation optimization method, which comprises the following steps: by combining chaotic mapping and a cuckoo algorithm, introducing a dynamic adaptive step factor and a dynamic discovery probability, and introducing disturbance reverse learning, constructing a multi-target optimization scheduling model; by improving the cuckoo optimization algorithm, performing population initialization by utilizing Logistic chaotic mapping, enhancing the randomness and global search ability of the algorithm, introducing a dynamic adaptive step factor and a dynamic discovery probability, increasing the diversity and flexibility of bird nest initialization, improving the global search ability, and introducing disturbance reverse learning, the algorithm has the advantages of improving the robustness of the algorithm, and improving the robustness of the algorithm. The algorithm performance is fully improved, and falling into a local optimal solution is avoided; compared with a common micro-grid, the improved comprehensive energy power grid has the advantages that the economic benefit can be effectively improved, pollutant emission is reduced, and the total cost of the whole system is reduced to a certain extent.

Description

Technical Field

[0001] The invention belongs to the technical field of energy system optimization, and in particular to a multi-objective operation optimization method for a combined cooling, heating and power integrated energy system. Background Art

[0002] With the rapid development of industry, traditional fossil resources are gradually exhausted, energy and environmental problems are becoming more serious, and the development and utilization of new energy sources such as wind and solar energy are gradually replacing traditional large-scale thermal power generation. However, the uncontrollability and intermittency of clean energy output constantly threaten the safe and stable operation of the power system, and unreasonable energy scheduling will cause serious wind and solar power abandonment. The scheduling of integrated energy systems is a complex, high-dimensional, nonlinear problem, and it is difficult to obtain an exact solution using traditional mathematical methods. Therefore, the scheduling problem of integrated energy power grids is mostly solved using intelligent algorithms to reduce the total operating cost of the system, reduce greenhouse gas and pollutant emissions, and improve the system's ability to absorb green energy;

[0003] The traditional cuckoo algorithm (CS) is a new heuristic algorithm that follows the Levy flight search mechanism by learning the reproduction strategy of cuckoos in nature. Its basic process is: first, randomly initialize NP nest locations, use Levy flight to generate new individuals, and retain the optimal solution; second, compare the random number with the discovery probability. If the random number is greater than the discovery probability, it means that the nest is found, and the inferior individuals are discarded according to the discovery probability; finally, generate the discarded number of new individuals according to the preference random walk method, and retain the optimal solution. Iterate in the above way until the global optimal solution is found.

[0004] Genetic Algorithm (GA) is an algorithm proposed by Professor Holland's team in the United States. This algorithm first initializes a population. This population consists of multiple individuals, each of which represents a potential solution to the problem. Parameters such as the size of the population and the encoding method of the individuals (such as binary, real number, etc.) need to be set according to the specific problem. The encoding method will affect each genetic operation of the algorithm. Common encoding methods include: binary encoding, decimal encoding, Gray encoding, etc. Then, the adaptive value of the individuals in the population is calculated. The fitness function is defined according to the goal of the problem, which measures the quality of each individual (i.e., the solution). By calculating the fitness value, we can understand the performance of each individual in solving the problem. In the selection stage, according to the individual fitness value, a part of the excellent individuals are selected to enter the next generation. This process simulates the principle of survival of the fittest in nature and retains the excellent genes in the population. The selection strategy is to pass on excellent genes to the next generation, and most of them are based on the fitness value method. From the steps, it can be seen that the selection probability is closely related to the fitness value, and the two are generally proportional, that is, the higher the fitness value, the greater the selection probability. Common selection strategies include roulette, random universal sampling, elite retention, tournaments, etc. The crossover phase is to simulate the gene recombination process in biological evolution. The selected individuals are randomly paired, and then some of their genes are exchanged with a certain crossover probability to produce new individuals. This process helps to increase the diversity of the population and may produce better solutions. After completing a round of selection, crossover, and mutation operations, determine whether the algorithm meets the termination conditions and whether the maximum population generation number is reached. If the termination conditions are met, the algorithm ends and outputs the optimal solution; otherwise, return to the fitness evaluation phase and continue the iterative search.

[0005] The principle of the particle swarm optimization algorithm (PSO) is to ignore the mass, volume and individual differences of the bird flock in the artificially created bird flock, abstract the bird flock into a particle swarm, and abstract the food into fitness. Use the information interaction between particles and between particles and groups. By tracking the optimal fitness of the group and the historical optimal fitness of the particle itself, the next moving direction is determined, and the optimal fitness is found through continuous iteration. The steps of the standard PSO algorithm are as follows: First, set the relevant PSO algorithm parameters, randomly generate the speed and position of several particles, and calculate the fitness. The current position and fitness are recorded as the individual optimal position, and the global leader is selected according to the principle of optimal fitness. Secondly, the position of all particles in the next step is calculated through the iterative formula and the fitness is updated. The current fitness is compared with the individual optimal fitness, and the better particle is selected as the individual optimal. After selecting the global leader, the above update link is repeated. Finally, the update is stopped when the convergence condition or the maximum number of iterations is reached, and the global optimal particle, i.e. the optimal solution, is output at the same time.

[0006] However, (1) although the cuckoo algorithm has the advantages of fewer parameter settings and faster convergence speed, in practical applications, the global search performance and local optimization accuracy of the algorithm need to be continuously improved. As a new type of intelligent optimization algorithm, the research on the cuckoo algorithm is still in its initial stage and may not be adaptable to specific problems. The cuckoo search algorithm uses a fixed step size for search, which limits the search ability for solutions to a certain extent. (2) Although the genetic algorithm has good global search capabilities, it is still easy to fall into local optimal solutions when facing problems in the field of microgrids. In addition, when facing large-scale microgrid systems, the genetic algorithm may not only require more computing resources, but also the convergence speed of the model is too slow. In microgrid scheduling problems with high dimensions and complex constraints, the genetic algorithm may not be able to achieve sufficiently high optimization accuracy. (3) Although the particle swarm algorithm is widely used as an intelligent algorithm, its particles are relatively concentrated in the later stages of iteration, and it is easy to fall into local optimal solutions, especially for objective functions with multiple extreme points. In addition, the particle swarm algorithm is relatively simple to implement, which also leads to low optimization accuracy.

[0007] To this end, technical personnel in this field proposed a multi-objective operation optimization method for a combined heat, cooling and power integrated energy system, aiming to establish a combined heat, cooling and power integrated energy system including major equipment such as wind power, photovoltaics, micro gas turbines, gas boilers, and electric refrigeration machines, and established a multi-objective optimization model with the economic and environmental benefits of the system as the goals. Summary of the invention

[0008] In order to solve the above technical problems, the present invention provides a multi-objective operation optimization method for a combined cooling, heating and power integrated energy system to solve the problems mentioned in the background technology.

[0009] A multi-objective operation optimization method for a combined cooling, heating and power energy system is proposed. By combining chaos mapping with the cuckoo algorithm, a dynamic adaptive step size factor and a dynamic discovery probability are introduced, and perturbation reverse learning is introduced to construct a multi-objective optimization scheduling model. The improved cuckoo optimization algorithm includes the following specific processes:

[0010] S1. Use Logistic chaotic mapping to initialize the population, that is, the variables required by the CCHP system, and set the cuckoo algorithm parameters, the number of bird nests, the dimension, and the maximum number of iterations;

[0011] S2, calculate the initial fitness value of each bird's nest and record the optimal solution;

[0012] S3, perform global search according to the dynamic step strategy, use Wright Flight to update the location of the current bird's nest, calculate the fitness value of the updated bird's nest location; compare it with the location before the update, and retain the optimal bird's nest;

[0013] S4, generate a random number r as the possibility of discovering foreign bird eggs, and compare it with the dynamic adaptive discovery probability Pa; if r>Pa, update the new bird nest position according to the discovery probability, compare the fitness values ​​of the bird nest positions before and after the update, and retain the bird nest position with the smaller fitness value; otherwise, retain the current bird nest and its fitness value;

[0014] S5, perform dimension-by-dimensional reverse learning on the selected updated solution, add a perturbation strategy, and update and save the optimal bird's nest and fitness value;

[0015] S6. Determine whether the current number of iterations has reached the maximum number of iterations. If the termination condition is met, exit the loop and output the global optimal result; otherwise, repeat S3 to continue iterating and optimizing.

[0016] Preferably, the Logistic chaos algorithm is used for initialization. The core of the Logistic chaos algorithm is the Logistic mapping function, which is expressed as:

[0017] x n+1 =μx n (1-x n )

[0018] Wherein, μ is the control parameter, satisfying μ∈(0,4], x∈(0,1). When the parameter is 3.57<μ≤4, it is in a completely chaotic state. In this interval, the motion trajectory of the equation is in a chaotic state. Here, μ=3.9999, x0=0.6.

[0019] Preferably, the search efficiency and convergence accuracy of the algorithm are improved by using a dynamically changing step size. The specific improvement is shown in the following formula:

[0020]

[0021] In the formula, iter represents the current number of iterations, max_iter represents the maximum number of iterations, and α max ,α min Represent the maximum and minimum values ​​of the step factor respectively.

[0022] Preferably, Pa is dynamically and adaptively adjusted, and the improved value of Pa shows a dynamic decreasing trend as the number of iterations increases, which is expressed as:

[0023]

[0024] In the early stage of the search, Pa is large, which enhances the global search capability; in the later stage of the search, Pa is small, which enhances the local search capability; the dynamic discovery probability Pa can adaptively switch between the local search and global search states with the number of iterations;

[0025] During the calculation process, the generated random number r is compared with the dynamic Pa. If r>Pa, the random walk phase is entered to randomly generate a new bird's nest position. The walk formula is expressed as:

[0026]

[0027] Among them, γ is the fitness factor that is uniformly distributed in the interval [0,1]. A random value representing a feasible solution of generation t.

[0028] Preferably, in the intelligent optimization algorithm, the iterative optimization process based on group search involves dynamically adjusting the search area to explore the solution space, introducing perturbation reverse learning, introducing the worst fitness value in each dimension, and improving the standard algorithm, using the current information of all bird nests, and improving the ability to avoid falling into the local optimum. The specific position update formula is:

[0029]

[0030] in, are the optimal fitness values ​​of the ith bird’s nest in the t+1th and tth dimensional iterations, respectively. are the upper and lower limits of the ith bird's nest in the tth dimension, is the worst fitness value of the i-th bird's nest in the t-th dimension update.

[0031] Preferably, in the multi-objective optimization scheduling model, the operating cost of the integrated energy system and the pollutant control cost are taken as the optimization targets, the economic cost is mainly composed of the interaction cost between the microgrid and the distribution network, the natural gas procurement cost and the unit operation and maintenance cost, and the investment cost of the system is determined, that is, the investment cost is not considered in the operation optimization, then the objective function is expressed as:

[0032] minC=min(C gird +C gas +C om +C en +C ab )

[0033] In the formula, C is the objective function value, C gird is the cost of electricity purchase and sale, C gas is the natural gas consumption cost, C om is the unit operation and maintenance cost, C en is the cost of pollutant treatment, C ab is the cost of wind and solar power abandonment; the specific expression is:

[0034]

[0035] C buy (t) = P buy (t)×cbuy (t)

[0036] C sell (t) = P sell (t)×c sell (t)

[0037] Where P buy (t), P sell (t) respectively represent the amount of electricity purchased and sold by the system to the large power grid at time t, c buy (t), c sell (t) respectively represent the price of electricity sold and purchased by the large power grid at time t;

[0038]

[0039] The gas turbine MT, internal combustion engine GE and gas boiler GB generate electricity by consuming natural gas, which will generate fuel costs in the process of power generation. The expression is as follows:

[0040]

[0041] In the formula, C OM (t) is the system operation and maintenance cost, K OM,n is the operation and maintenance coefficient of the nth type of equipment, P n (t) represents the output power of the nth device, and N represents the total number of devices;

[0042] Generators generate pollutants during the power generation process. The cost of pollutant emission control is expressed as:

[0043]

[0044] k is the type number of the pollutant emitted, K is the total number of pollutant types, α ik is the emission coefficient of k pollutants of different types of units, α gk is the emission coefficient of pollutant k generated by the system, β k The cost required to treat pollutant k;

[0045]

[0046] In the formula, λ1 and λ2 represent the cost coefficients of wind power abandonment and solar power abandonment. Indicates the maximum wind and solar power output at time t in the cycle, P WT (t), P PV (t) represents the actual wind and solar power output at time t.

[0047] Preferably, in constructing a multi-objective optimization scheduling model, the system constraints include electric power balance constraints, thermal power balance constraints, cooling power balance constraints, and energy storage constraints; wherein the electric power balance constraint is expressed as:

[0048] P WT (t)+P PV (t)+P grid (t)+P ess (t)+P MT (t)+P GE (t) = P EH (t)+L e (t)

[0049] Where P WT (t) represents the output power of wind power generation at time t, P PV (t) represents the output power of photovoltaic power generation at time t; P ess (t) represents the output power of the battery at time t. ess When (t)>0, it means the battery is discharged. At this time, P ess (t) = P dis (t), when P ess When (t) < 0, it means the battery is charging. ess (t) = P cha (t); P grid (t) represents the interaction power between the integrated energy microgrid and the large power grid. grid When (t)>0, it means that the microgrid buys electricity from the large power grid, P grid (t) = P buy (t), on the contrary, sell electricity to the large power grid, at this time P grid (t) = P sell (t); L e (t) is the load power consumption at time t;

[0050] The thermal power balance constraint is expressed as:

[0051] P GB (t)+P WH (t)-P ac_h (t) = L h (t)+P ess,h (t)

[0052] Where P ess,h (t) represents the thermal power output of the heat storage tank at time t. If it is greater than zero, it means heat release, otherwise it means heat absorption; L h (t) represents the heat power of the load at time t;

[0053] The cold power balance constraint is expressed as:

[0054] Pec,c (t)+P ac (t) = L c (t)+P ess,c (t)

[0055] Where P ess,c (t) represents the cooling power output by the cold storage device at time t. If it is greater than zero, it means cooling, otherwise it means cooling; L c (t) represents the cooling power used by the load at time t;

[0056] The energy storage constraint is expressed as:

[0057]

[0058] Where n represents the nth energy storage device; E bat,n They represent the capacity of the nth type of energy storage respectively; are binary variables, representing the charging and discharging states of energy storage, Indicates that it is in charging state, otherwise it is in discharging state.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. The present invention proposes an improved cuckoo CDCS optimization algorithm based on a combination of chaotic mapping and cuckoo algorithm. Aiming at the slow convergence speed of the traditional cuckoo algorithm and the limited search capability of the microgrid system with a complex constraint environment, the present invention uses Logistic chaotic mapping for population initialization to enhance the randomness and global search capability of the algorithm, introduces a dynamic adaptive step size factor and a dynamic discovery probability, increases the diversity and flexibility of bird's nest initialization, and improves the global search capability; introduces perturbation reverse learning to fully improve the algorithm performance and avoid falling into the local optimal solution.

[0061] 2. The present invention can effectively improve economic benefits and reduce pollutant emissions by comparing the improved integrated energy grid with the ordinary microgrid.

[0062] 3. Compared with other algorithms, the improved cuckoo algorithm of the present invention has better performance in solving the model and can reduce the total cost of the entire system to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a flow chart of the improved cuckoo algorithm of the present invention;

[0064] Figure 2 This is the Logistic chaos simulation effect diagram of the present invention;

[0065] Figure 3 This is the optimization result diagram of the typical summer day scene 1 of the present invention;

[0066] Figure 4 This is the optimization result diagram of the typical summer day scene 2 of the present invention;

[0067] Figure 5 This is the optimization result diagram of the typical summer day scene 3 of the present invention;

[0068] Figure 6 This is the optimization result diagram of the winter typical day scene 1 of the present invention;

[0069] Figure 7 This is the optimization result diagram of the winter typical day scene 2 of the present invention;

[0070] Figure 8 This is the optimization result diagram of the winter typical day scene 3 of the present invention;

[0071] Fig. 9 The optimization result graph of the present invention using the CS and CDCS algorithms;

[0072] Fig.10 These are the fitting curves of the CDCS, CS and PSO algorithms of the present invention. DETAILED DESCRIPTION

[0073] 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.

[0074] As attached Figure 1 To Attachment Fig.10 As shown:

[0075] Embodiment 1: The present invention provides a multi-objective operation optimization method for a combined cooling, heating and power integrated energy system. By combining chaos mapping with a cuckoo algorithm, a dynamic adaptive step size factor and a dynamic discovery probability are introduced, and perturbation reverse learning is introduced to construct a multi-objective optimization scheduling model. The improved cuckoo optimization algorithm includes the following specific processes:

[0076] S1. Use Logistic chaotic mapping to initialize the population, that is, the variables required by the CCHP system, and set the cuckoo algorithm parameters, the number of bird nests, the dimension, and the maximum number of iterations;

[0077] S2, calculate the initial fitness value of each bird's nest and record the optimal solution;

[0078] S3, perform global search according to the dynamic step strategy, use Wright Flight to update the location of the current bird's nest, calculate the fitness value of the updated bird's nest location; compare it with the location before the update, and retain the optimal bird's nest;

[0079] S4, generate a random number r as the possibility of discovering foreign bird eggs, and compare it with the dynamic adaptive discovery probability Pa; if r>Pa, update the new bird nest position according to the discovery probability, compare the fitness values ​​of the bird nest positions before and after the update, and retain the bird nest position with the smaller fitness value; otherwise, retain the current bird nest and its fitness value;

[0080] S5, perform dimension-by-dimensional reverse learning on the selected updated solution, add a perturbation strategy, and update and save the optimal bird's nest and fitness value;

[0081] S6. Determine whether the current number of iterations has reached the maximum number of iterations. If the termination condition is met, exit the loop and output the global optimal result; otherwise, repeat S3 to continue iterating and optimizing.

[0082] The Logistic chaos algorithm is used for initialization. The core of the Logistic chaos algorithm is the Logistic mapping function, which is expressed as:

[0083] x n+1 =μx n (1-x n )

[0084] Wherein, μ is the control parameter, satisfying μ∈(0,4], x∈(0,1). When the parameter is 3.57<μ≤4, it is in a completely chaotic state. In this interval, the motion trajectory of the equation is in a chaotic state. Here, μ=3.9999, x0=0.6.

[0085] In the target update formula, α is the step size control factor, and the standard cuckoo algorithm generally takes α = 0.01. However, the fixed step size setting limits the search for solutions to a certain extent. In the early stage of the algorithm, a larger step size can expand the search space and achieve a better global search; as the number of iterations increases, shortening the step size in the later stage of the algorithm can be more conducive to local search and improve search accuracy; therefore, the search efficiency and convergence accuracy of the algorithm are improved by using a dynamically changing step size. The specific improvement is shown in the following formula:

[0086]

[0087] In the formula, iter represents the current number of iterations, max_iter represents the maximum number of iterations, and α max ,α min Represent the maximum and minimum values ​​of the step factor respectively.

[0088] The probability of the host discovering cuckoo eggs is Pa, which indicates the probability of the host bird identifying foreign eggs in the nest. Most researchers set Pa to a fixed value of 0.25, but the cuckoo algorithm uses a fixed discovery probability Pa, which is not conducive to the local search and global search of the dynamic equilibrium algorithm. Therefore, Pa is dynamically and adaptively adjusted. After the improvement, the value of Pa shows a dynamic decreasing trend with the increase of the number of iterations, which is expressed as:

[0089]

[0090] In the early stage of the search, Pa is large, which enhances the global search capability; in the later stage of the search, Pa is small, which enhances the local search capability; the dynamic discovery probability Pa can adaptively switch between the local search and global search states with the number of iterations;

[0091] During the calculation process, the generated random number r is compared with the dynamic Pa. If r>Pa, the random walk phase is entered to randomly generate a new bird's nest position. The walk formula is expressed as:

[0092]

[0093] Among them, γ is the fitness factor that is uniformly distributed in the interval [0,1]. A random value representing a feasible solution of generation t.

[0094] In the intelligent optimization algorithm, the iterative optimization process based on group search involves dynamically adjusting the search area to explore the solution space. By introducing perturbation reverse learning, the worst fitness value in each dimension is introduced, and the standard algorithm is improved to use the current information of all bird nests to improve the ability to avoid falling into the local optimum. The specific position update formula is:

[0095]

[0096] in, are the optimal fitness values ​​of the ith bird’s nest in the t+1th and tth dimensional iterations, respectively. are the upper and lower limits of the ith bird's nest in the tth dimension, is the worst fitness value of the i-th bird's nest in the t-th dimension update.

[0097] In the multi-objective optimization scheduling model, the operating cost of the integrated energy system and the cost of pollutant control are taken as the optimization targets. The economic cost is mainly composed of the interaction cost between the microgrid and the distribution network, the natural gas procurement cost, and the unit operation and maintenance cost. The investment cost of the system is determined, that is, the investment cost is not considered in the operation optimization. The objective function is expressed as:

[0098] minC=min(C gird +C gas +C om +Cen +C ab )

[0099] In the formula, C is the objective function value, C gird is the cost of electricity purchase and sale, C gas is the natural gas consumption cost, C om is the unit operation and maintenance cost, C en is the cost of pollutant treatment, C ab is the cost of wind and solar power abandonment; the specific expression is:

[0100]

[0101] C buy (t) = P buy (t)×c buy (t)

[0102] C sell (t) = P sell (t)×c sell (t)

[0103] Where P buy (t), P sell (t) respectively represent the amount of electricity purchased and sold by the system to the large power grid at time t, c buy (t), c sell (t) respectively represent the price of electricity sold and purchased by the large power grid at time t;

[0104]

[0105] The gas turbine MT, internal combustion engine GE and gas boiler GB generate electricity by consuming natural gas, which will generate fuel costs in the process of power generation. The expression is as follows:

[0106]

[0107] In the formula, C OM (t) is the system operation and maintenance cost, K OM,n is the operation and maintenance coefficient of the nth type of equipment, P n (t) represents the output power of the nth device, and N represents the total number of devices;

[0108] Generators generate pollutants during the power generation process. The cost of pollutant emission control is expressed as:

[0109]

[0110] k is the type number of the pollutant emitted, K is the total number of pollutant types, α ik is the emission coefficient of k pollutants of different types of units, α gkis the emission coefficient of pollutant k generated by the system, β k The cost required to treat pollutant k;

[0111]

[0112] In the formula, λ1 and λ2 represent the cost coefficients of wind power abandonment and solar power abandonment. Indicates the maximum wind and solar power output at time t in the cycle, P WT (t), P PV (t) represents the actual wind and solar power output at time t.

[0113] In constructing the multi-objective optimization scheduling model, the system constraints include electric power balance constraint, thermal power balance constraint, cooling power balance constraint, and energy storage constraint; the electric power balance constraint is expressed as:

[0114] P WT (t)+P PV (t)+P grid (t)+P ess (t)+P MT (t)+P GE (t) = P EH (t)+L e (t)

[0115] Where P WT (t) represents the output power of wind power generation at time t, P PV (t) represents the output power of photovoltaic power generation at time t; P ess (t) represents the output power of the battery at time t. ess When (t)>0, it means the battery is discharged. At this time, P ess (t) = P dis (t), when P ess When (t) < 0, it means the battery is charging. ess (t) = P cha (t); P grid (t) represents the interaction power between the integrated energy microgrid and the large power grid. grid When (t)>0, it means that the microgrid buys electricity from the large power grid, P grid (t) = P buy (t), on the contrary, sell electricity to the large power grid, at this time P grid (t) = P sell (t); L e (t) is the load power consumption at time t;

[0116] The thermal power balance constraint is expressed as:

[0117] P GB (t)+PWH (t)-P ac_h (t) = L h (t)+P ess,h (t)

[0118] Where P ess,h (t) represents the thermal power output of the heat storage tank at time t. If it is greater than zero, it means heat release, otherwise it means heat absorption; L h (t) represents the heat power of the load at time t;

[0119] The cold power balance constraint is expressed as:

[0120] P ec,c (t)+P ac (t) = L c (t)+P ess,c (t)

[0121] Where P ess,c (t) represents the cooling power output by the cold storage device at time t. If it is greater than zero, it means cooling, otherwise it means cooling; L c (t) represents the cooling power used by the load at time t;

[0122] The energy storage constraint is expressed as:

[0123]

[0124] Where n represents the nth energy storage device; E bat,n They represent the capacity of the nth type of energy storage respectively; are binary variables, representing the charging and discharging states of energy storage, Indicates that it is in charging state, otherwise it is in discharging state.

[0125] From the above, it can be seen that the present invention uses Logistic chaotic mapping to initialize the population, enhances the randomness and global search ability of the algorithm, introduces a dynamic adaptive step size factor and a dynamic discovery probability, increases the diversity and flexibility of bird nest initialization, and improves the global search ability; introduces perturbation reverse learning, fully improves the algorithm performance, and avoids falling into the local optimal solution.

[0126] Basic data description:

[0127] The present invention uses matlab software to solve the optimization scheduling model. Pollutant emissions and treatment costs are shown in Table 1. The unit price of natural gas in the system is 3.12 yuan / M3, and the time division and purchase and sale electricity prices are shown in Table 2.

[0128] Table 1 Pollutant treatment costs

[0129]

[0130] Table 2 Time division and electricity purchase and sale prices

[0131]

[0132] The relevant parameters of each unit in the integrated energy microgrid are shown in Tables 3 and 4 below.

[0133] Table 3 Unit operation and maintenance costs

[0134]

[0135] Table 4 Distributed power supply parameters

[0136]

[0137] In order to compare the advantages of the combined cooling, heating and power model and the impact of environmental pollution costs on the results, the following three scenarios are selected as references, which are recorded in the present invention as follows:

[0138] Scenario 1: Combined cooling, heating and power, taking environmental costs into consideration.

[0139] Scenario 2: Separate production of cooling, heating and electricity, and take environmental costs into consideration.

[0140] In this case, the electric energy is mainly met by the external large power grid, clean energy and power generation equipment, the cold energy is mainly met by electric refrigerators, and the heat energy is mainly met by gas boilers GB.

[0141] Scenario 3: Considering cogeneration of heat, cooling and power, but ignoring environmental costs.

[0142] Scenario 1 is based on Scenario 2. On the supply side, energy conversion equipment is used to improve the system's multi-energy supply capacity. Thermal energy is provided by GB and waste heat recovery, and cold energy is provided by electric refrigeration and absorption chillers.

[0143] Finally, based on scenario 1, the improvement of the CDCS algorithm proposed in the present invention compared with the PSO and CS algorithms is demonstrated.

[0144] Analysis of operation results in different scenarios

[0145] Table 5 Optimization results

[0146]

[0147] In order to compare the stability of the algorithms and eliminate the errors caused by unpredictable methods, the scenario 1 model was run 30 times using the CS and CDCS algorithms. The box plots show the results of the two algorithms, as shown in Fig. 9 As shown. Using the Goldstein-Price function, CDCS, CS and PSO algorithms were compared and tested to determine which algorithm is more effective, and the fitness curve was iteratively generated. The results are shown in Fig.10As shown. Fig.10 As shown, the CDCS algorithm can find the optimal solution faster and more effectively than the CS and PSO algorithms. The results fully demonstrate the superiority of the CDCS algorithm in solving optimization problems and provide strong support for the practicality of the algorithm in this work.

[0148] Analysis of optimization results for a typical summer day

[0149] Scenario 1 Analysis

[0150] It can be seen from Table 5 that the comprehensive economic cost of the system is 2226.29 yuan, and the pollutant emissions are 1394.85g. Figure 3 The electric power and thermal power optimization results of scenario 1 are shown in the figure. Due to the comprehensive consideration of economic and environmental indicators, in order to obtain less pollutant emissions, batteries and power purchases from the power grid are not conducive to environmental indicators, so they are less involved. Wind power and photovoltaic power generation are clean energy, so wind power and photovoltaic output are given priority during the entire operation process, and are basically in maximum power tracking. Figure 3 As shown in a, at 23-6, since the electricity purchase price is low at this time, the cost of purchasing electricity is lower than the cost of generating electricity with the micro gas turbine, so the system purchases electricity from the grid during this period. Figure 3 b. At this time, the output of the electric refrigerator is higher than that of the absorption refrigerator. At 6-11 and 17-23, the electricity price is at the peak value. The power generation cost of the micro gas turbine is lower than the purchase price of electricity from the large power grid, and the environmental pollution of gas turbine power generation is less than that of power grid purchase. Therefore, the power supply is mainly provided by gas turbines, and the power grid purchase is used as a supplement. Figure 3 b Cooling power balance diagram. Since the greater the gas turbine power, the more waste heat is generated, the absorption chiller will supply a large amount of cooling load during this period, and the electric chiller will supplement it. When 11-17, the cooling load is at its peak, while the electric load is at its low point. Figure 3 a. The output of the absorption refrigerator is insufficient, so the output of the electric refrigerator is greater.

[0151] observe Figure 4 Scenario 2 does not consider the cogeneration of cooling, heating and power, so the waste heat of the gas turbine cannot be fully utilized. The cooling load is supplied by the electric refrigerator, and the total power load of the system increases. However, due to the consideration of environmental costs and the output limit of the unit, the power purchase amount of the power grid increases. In summary, the total cost of scenario 2 increased by 15.11% to 2562.63 yuan, the pollutant emissions were 1462.01, and the cost was 434.63 yuan.

[0152] The optimization results of scenario 3 are as follows: Figure 5By comparing Scenario 1 and Scenario 3, since environmental factors are not taken into account, and the electricity price will be significantly lower than the unit price of natural gas, when the electricity price is not at its peak, more power grid purchases are used to meet the cooling load and electricity load requirements. On the surface, the total system cost is reduced to 1813.83 yuan. However, the pollutant emissions in Scenario 3 increased by 7.65% compared with Scenario 1, reaching 1501.57g, and the treatment cost reached 502.09 yuan, so the total cost was 2315.92 yuan, which was higher than the cost in Scenario 1. In summary, although Scenario 3 reduces the total cost on the surface, the corresponding environmental pollution increases.

[0153] Analysis of optimization results for a typical winter day

[0154] Scenario 1 Analysis

[0155] From the optimization results in Table 5, the total cost of the system is 2195.63 yuan, and the pollutant emissions are 1237.41g. The optimization results of electrical power and thermal power are as follows Figure 6 As shown in the figure, throughout the day, since wind turbines and photovoltaics are clean energy, have no pollutant emissions, and have low operation and maintenance costs, wind turbines and photovoltaics are used for power generation, and micro gas turbines are used as supplements. In addition, at 9-13 and 19-22, the electricity price is at its peak, and the excess power generation can be sold to reduce its own costs. At 9-12 and 16-22, since the electricity load is large at this time, and the pollution emissions of power grid purchases are significantly higher than gas turbines, gas turbines are mainly used to meet the electricity load demand at this time. At the same time, waste heat boilers make full use of the waste heat of gas turbines for heating, and the insufficient part is supplemented by gas boilers. In the period from 23 to 8, the electricity load is at a low value period. In addition to wind power output, gas turbines are used to supplement, but the heat load is at its peak at this time, and the waste heat supply of gas turbines is insufficient, so the output of gas boilers is higher than that of waste heat boilers.

[0156] observe Figure 7 By comparing Scenario 1 and Scenario 2, it is found that since Scenario 2 is a heat and power generation, the waste heat generated by the gas turbine power generation is not fully utilized. For the heat load, it can only rely on gas boilers to consume natural gas to generate heat, so the gas purchase volume is increased, and the total system cost is 30.43% higher than Scenario 1, which is 2,863.67 yuan. Figure 8 As is known, in electrical power optimization, the output of the gas turbine is relatively reduced.

[0157] By comparing Scenario 1 and Scenario 3, we found that because Scenario 3 did not consider the environmental pollution cost, the total cost of the system was reduced on the surface, which was 1873.79 yuan. However, the pollutant emissions were significantly higher than Scenario 1, increasing by 8.39% to 1341.17g, and the treatment cost was 354.83 yuan. Therefore, the total cost of Scenario 3 was 2228.62, which was higher than Scenario 1. Figure 6 a and Figure 8 The power optimization results of a show that in the low electricity price period, the results of scenario 3 are more likely to be purchased from the power grid to meet the power load demand, so the system cost is reduced, but the pollutants are significantly increased. Figure 6 b and Figure 8 b. Thermal power optimization requirements. Since the gas turbine output is reduced, the heat production of the waste heat boiler is reduced and the gas boiler output is increased.

[0158] The running results of the three algorithms in scenario 1 are shown in Table 5. Considering the economy and environmental protection of the system, the PSO algorithm, CS algorithm and the CDCS algorithm proposed in the present invention are used to solve the model, and the solution results are shown below. As can be seen from Table 5, the final verification result is that when considering the comprehensive cost, on a typical summer day, the improved cuckoo algorithm is 10.68% and 1.13% lower than PSO and CS respectively; on a typical winter day, the improved cuckoo algorithm is 11.87% and 0.25% lower than PSO and CS respectively. Moreover, whether it is a model of a typical summer or winter day, the improved cuckoo algorithm is also better than the CS algorithm and the PSO algorithm in terms of environmental protection.

[0159] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0160] It should be understood that in the development of any actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, without undue experimentation, the development effort will be a routine task of design, fabrication, and production.

[0161] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A multi-objective operation optimization method for a combined cooling, heating and power integrated energy system, characterized by: By combining chaos mapping with the cuckoo algorithm, introducing dynamic adaptive step size factors and dynamic discovery probability, and introducing perturbation reverse learning, a multi-objective optimization scheduling model is constructed. The improved cuckoo optimization algorithm includes the following specific processes: S1. Use Logistic chaotic mapping to initialize the population, that is, the variables required by the CCHP system, and set the cuckoo algorithm parameters, the number of bird nests, the dimension, and the maximum number of iterations; S2, calculate the initial fitness value of each bird's nest and record the optimal solution; S3, perform global search according to the dynamic step strategy, use Wright Flight to update the location of the current bird's nest, calculate the fitness value of the updated bird's nest location; compare it with the location before the update, and retain the optimal bird's nest; S4, generate a random number r as the possibility of discovering foreign bird eggs, and compare it with the dynamic adaptive discovery probability Pa; if r>Pa, update the new bird nest position according to the discovery probability, compare the fitness values ​​of the bird nest positions before and after the update, and retain the bird nest position with the smaller fitness value; otherwise, retain the current bird nest and its fitness value; S5, perform dimension-by-dimensional reverse learning on the selected updated solution, add a perturbation strategy, and update and save the optimal bird's nest and fitness value; S6. Determine whether the current number of iterations has reached the maximum number of iterations. If the termination condition is met, exit the loop and output the global optimal result; otherwise, repeat S3 to continue iterating and optimizing.

2. A multi-objective operation optimization method for a combined cooling, heating and power integrated energy system as claimed in claim 1, characterized in that: The Logistic chaos algorithm is used for initialization. The core of the Logistic chaos algorithm is the Logistic mapping function, which is expressed as: x n+1 =μx n (1-x n ) Wherein, μ is the control parameter, satisfying μ∈(0,4], x∈(0,1). When the parameter is 3.57<μ≤4, it is in a completely chaotic state. In this interval, the motion trajectory of the equation is in a chaotic state. Here, μ=3.9999, x0=0.

6.

3. A multi-objective operation optimization method for a combined cooling, heating and power integrated energy system as claimed in claim 1, characterized in that: The search efficiency and convergence accuracy of the algorithm are improved by using a dynamically changing step size. The specific improvements are shown in the following formula: In the formula, iter represents the current number of iterations, max_iter represents the maximum number of iterations, and α max ,α min Represent the maximum and minimum values ​​of the step factor respectively.

4. A multi-objective operation optimization method for a combined cooling, heating and power integrated energy system as claimed in claim 1, characterized in that: Pa is dynamically and adaptively adjusted. After improvement, the value of Pa shows a dynamic decreasing trend as the number of iterations increases, which can be expressed as: In the early stage of the search, Pa is large, which enhances the global search capability; in the later stage of the search, Pa is small, which enhances the local search capability; the dynamic discovery probability Pa can adaptively switch between the local search and global search states with the number of iterations; During the calculation process, the generated random number r is compared with the dynamic Pa. If r>Pa, the random walk phase is entered to randomly generate a new bird's nest position. The walk formula is expressed as: Among them, γ is the fitness factor that is uniformly distributed in the interval [0,1]. A random value representing a feasible solution of generation t.

5. The multi-objective operation optimization method of a combined cooling, heating and power integrated energy system according to claim 1, characterized in that: In the intelligent optimization algorithm, the iterative optimization process based on group search involves dynamically adjusting the search area to explore the solution space. By introducing perturbation reverse learning, the worst fitness value in each dimension is introduced, and the standard algorithm is improved to use the current information of all bird nests to improve the ability to avoid falling into the local optimum. The specific position update formula is: in, are the optimal fitness values ​​of the ith bird’s nest in the t+1th and tth dimensional iterations, respectively. are the upper and lower limits of the ith bird's nest in the tth dimension, is the worst fitness value of the i-th bird's nest in the t-th dimension update.

6. A multi-objective operation optimization method for a combined cooling, heating and power integrated energy system as claimed in claim 1, characterized in that: In the multi-objective optimization scheduling model, the operating cost of the integrated energy system and the cost of pollutant control are taken as the optimization targets. The economic cost is mainly composed of the interaction cost between the microgrid and the distribution network, the natural gas procurement cost, and the unit operation and maintenance cost. The investment cost of the system is determined, that is, the investment cost is not considered in the operation optimization. The objective function is expressed as: minC=min(C gird +C gas +C om +C en +C ab ) In the formula, C is the objective function value, C gird is the cost of electricity purchase and sale, C gas is the natural gas consumption cost, C om is the unit operation and maintenance cost, C en is the cost of pollutant treatment, C ab is the cost of wind and solar power abandonment; the specific expression is: C buy (t)=P buy (t)×c buy (t) C sell (t)=P sell (t)×c sell (t) Where P buy (t), P sell (t) respectively represent the amount of electricity purchased and sold by the system to the large power grid at time t, c buy (t), c sell (t) respectively represent the price of electricity sold and purchased by the large power grid at time t; The gas turbine MT, internal combustion engine GE and gas boiler GB generate electricity by consuming natural gas, which will generate fuel costs in the process of power generation. The expression is as follows: In the formula, C OM (t) is the system operation and maintenance cost, K OM,n is the operation and maintenance coefficient of the nth type of equipment, P n (t) represents the output power of the nth device, and N represents the total number of devices; Generators generate pollutants during the power generation process. The cost of pollutant emission control is expressed as: k is the type number of the pollutant emitted, K is the total number of pollutant types, α ik is the emission coefficient of k pollutants of different types of units, α gk is the emission coefficient of pollutant k generated by the system, β k The cost required to treat pollutant k; In the formula, λ1 and λ2 represent the cost coefficients of wind power abandonment and solar power abandonment. Indicates the maximum wind and solar power output at time t in the cycle, P WT (t), P PV (t) represents the actual wind and solar power output at time t.

7. A multi-objective operation optimization method for a combined cooling, heating and power integrated energy system as claimed in claim 6, characterized in that: In constructing the multi-objective optimization scheduling model, the system constraints include electric power balance constraint, thermal power balance constraint, cooling power balance constraint, and energy storage constraint; the electric power balance constraint is expressed as: P WT (t)+P PV (t)+P grid (t)+P ess (t)+P MT (t)+P GE (t)=P EH (t)+L e (t) Where P WT (t) represents the output power of wind power generation at time t, P PV (t) represents the output power of photovoltaic power generation at time t; P ess (t) represents the output power of the battery at time t. ess When (t)>0, it means the battery is discharged. At this time, P ess (t) = P dis (t), when P ess When (t) < 0, it means the battery is charging. ess (t) = P cha (t); P grid (t) represents the interaction power between the integrated energy microgrid and the large power grid. grid When (t)>0, it means that the microgrid buys electricity from the large power grid, P grid (t) = P buy (t), on the contrary, sell electricity to the large power grid, at this time P grid (t) = P sell (t); L e (t) is the load power consumption at time t; The thermal power balance constraint is expressed as: P GB (t)+P WH (t)-P ac_h (t)=L h (t)+P ess,h (t) Where P ess,h (t) represents the thermal power output of the heat storage tank at time t. If it is greater than zero, it means heat release, otherwise it means heat absorption; L h (t) represents the heat power of the load at time t; The cold power balance constraint is expressed as: P ec,c (t)+P ac (t)=L c (t)+P ess,c (t) Where P ess,c (t) represents the cooling power output by the cold storage device at time t. If it is greater than zero, it means cooling, otherwise it means cooling; L c (t) represents the cooling power used by the load at time t; The energy storage constraint is expressed as: Where n represents the nth energy storage device; E bat,n They represent the capacity of the nth type of energy storage respectively; are binary variables, representing the charging and discharging states of energy storage, Indicates that it is in charging state, otherwise it is in discharging state.

Images