A method, device, medium and product for generating the installed capacity domain of a power system

Abstract

The present application discloses a method, device, medium and product for generating the installed capacity domain of a power system, relating to the field of generating the installed capacity domain of a power system. The method includes obtaining the expected upper limit value of the one-time investment coefficient input and the expected lower limit value of the new energy consumption rate of the target power system, and using the installed capacity domain generation model of the power system to determine the installed capacity domain of the target power system; the installed capacity domain is composed of the upper bound of the installed capacity domain and the lower bound of the installed capacity domain; wherein, the installed capacity domain generation model of the power system is obtained by training the CNN-BiLSTM-BiGRU neural network model using a training data set; the training data set is established using a multi-objective coronavirus disease optimization algorithm and a wave search algorithm; the CNN-BiLSTM-BiGRU neural network model includes a convolutional neural network, a bidirectional long short-term memory neural network and a bidirectional gated recurrent unit connected in sequence. The present application improves the accuracy of generating the installed capacity domain of a power system.

Description

Technical Field

[0001] The present application relates to the field of power system installed capacity domain generation, and in particular to a power system installed capacity domain generation method, equipment, medium and product. Background Art

[0002] With the deepening of global energy transformation, the proportion of renewable energy electricity has increased year by year. Due to the randomness and volatility of renewable energy such as wind power and photovoltaics, the current power system is facing the problem of high new energy absorption rate and safe and stable operation of the power system. The new energy absorption rate has also become an important assessment indicator of the current power system. Reasonable power structure arrangement on the source side is a key factor affecting the new energy absorption rate. Excessive construction of new energy unit capacity or too little energy storage configuration will lead to a large amount of new energy abandonment. Even if appropriate operation and scheduling strategies can promote the absorption of some new energy, the increase in new energy absorption rate is limited due to the safety requirements of system operation. At the same time, the existing capacity design method is difficult to fully describe the multi-dimensional attributes of power system capacity, and lacks intuitive decision support tools when dealing with the development trend of power supply capacity. The installed capacity domain of the power system constructed based on domain theory can clearly show the capacity boundary of the power system and provide capacity boundary support for the power supply development of the power system from a long-term perspective. Summary of the invention

[0003] The purpose of this application is to provide a method, device, medium and product for generating an installed capacity domain of an electric power system, so as to improve the accuracy of generating the installed capacity domain of the electric power system.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a method for generating an installed capacity domain of a power system, comprising:

[0006] Obtain the expected upper limit value of the one-time investment coefficient of the target power system and the expected lower limit value of the new energy consumption rate;

[0007] According to the upper limit expected value of the one-time investment coefficient and the lower limit expected value of the new energy consumption rate, the installed capacity domain of the target power system is determined by using the installed capacity domain generation model of the power system; the installed capacity domain consists of an upper bound of the installed capacity domain and a lower bound of the installed capacity domain; wherein, the installed capacity domain generation model of the power system is obtained by training a CNN-BiLSTM-BiGRU neural network model using a training data set; the training data set is established by using a multi-objective coronavirus disease optimization algorithm and a wave search algorithm; the training data set includes the upper limit of the one-time investment coefficient input, the lower limit of the new energy consumption rate, and the corresponding installed capacity domain labels at each stage of the training power system; the CNN-BiLSTM-BiGRU neural network model includes a convolutional neural network, a bidirectional long short-term memory neural network, and a bidirectional gated recurrent unit connected in sequence.

[0008] Optionally, training the CNN-BiLSTM-BiGRU neural network model using the training data set specifically includes:

[0009] Construct a training data set;

[0010] Input the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate at each stage of the training power system into the current CNN-BiLSTM-BiGRU neural network model to obtain the predicted value of the upper bound of the installed capacity domain and the predicted value of the lower bound of the installed capacity domain;

[0011] Determine the loss function value according to the predicted value of the upper bound of the installed capacity domain, the predicted value of the lower bound of the installed capacity domain, and the corresponding installed capacity domain label;

[0012] Judge whether the training end condition is satisfied; the training end condition is reaching the maximum number of iterations or the loss function value is less than a preset value;

[0013] If so, end the training and use the current CNN-BiLSTM-BiGRU neural network model as the installed capacity domain generation model of the power system;

[0014] If not, adjust the model parameters of the current CNN-BiLSTM-BiGRU neural network model according to the loss function value and return "Input the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate at each stage of the training power system into the current CNN-BiLSTM-BiGRU neural network model to obtain the predicted value of the upper bound of the installed capacity domain and the predicted value of the lower bound of the installed capacity domain".

[0015] Optionally, constructing a training data set specifically includes:

[0016] Use a random function to generate the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate at each stage of the training power system;

[0017] Generate the initial installed capacity domain of the power system for training by using the multi-objective coronavirus disease optimization algorithm;

[0018] Based on the initial installed capacity domain, use the wave search algorithm to determine the installed capacity domain labels corresponding to the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate for each stage of the power system for training, and obtain the training data set.

[0019] Optionally, generating the initial installed capacity domain of the power system for training by using the multi-objective coronavirus disease optimization algorithm specifically includes:

[0020] Establish a multi-objective optimization model; the multi-objective optimization model takes the minimum total coefficient and the minimum new energy consumption rate as the objective functions, and takes the unit capacity constraint of the unit, the unit combination constraint, and the clean and low-carbon driving constraint as the constraint conditions;

[0021] Use the multi-objective coronavirus disease optimization algorithm to solve the multi-objective optimization model and determine the optimal solution set of the installed capacity domain;

[0022] According to the optimal solution set of the installed capacity domain, determine the initial installed capacity domain of the power system for training; the upper bound of the initial installed capacity domain is the maximum value of the power source capacity in the optimal solution set of the installed capacity domain; the lower bound of the initial installed capacity domain is the minimum value of the power source capacity in the optimal solution set of the installed capacity domain.

[0023] Optionally, the objective function is:

[0024]

[0025] where F a is the total cost; T a is the number of days in the current stage; is the unit capacity investment cost of the power source equipment in the nth stage; is the installed capacity of the power source equipment in the nth stage; is the installed capacity of the power source equipment in the (n + 1)th stage; is the unit capacity maintenance cost of the power source equipment in the nth stage; τ s is the weight coefficient of the s-th scenario; is the unit power operation cost of the electrochemical energy storage equipment in the nth stage; is the charging power of the electrochemical energy storage equipment in the nth stage, the s-th scenario, and the t-th time period; is the discharging power of the electrochemical energy storage equipment in the nth stage, the s-th scenario, and the t-th time period; Δt is the time interval; is the unit power operation cost of thermal power in the nth stage; is the actual output of the thermal power at the nth stage, s-th scenario, and t-th time period; is the unit power penalty cost for load shedding, is the actual load shedding at the nth stage, s-th scenario, and t-th time period; G a is the new energy consumption rate, N r is the number of stages for power capacity decision-making; is the actual output of the wind power at the nth stage, s-th scenario, and t-th time period; is the actual output of the photovoltaic power at the nth stage, s-th scenario, and t-th time period; is the installed capacity of the wind power at the nth stage; is the installed capacity of the photovoltaic power at the nth stage; is the predicted capacity factor of the wind power at the nth stage, s-th scenario, and t-th time period; is the predicted capacity factor of the photovoltaic power at the nth stage, s-th scenario, and t-th time period.

[0026] Optionally, the unit capacity constraint is:

[0027]

[0028] wherein, is the installed capacity of the power supply equipment at the nth stage; is the installed capacity of the power supply equipment at the (n + 1)-th stage; is the lower limit of the installed capacity at the nth stage; is the upper limit of the installed capacity at the nth stage; is the unit capacity investment cost of the power supply equipment at the nth stage, is the installed capacity of the power supply equipment at the (n - 1)-th stage; is the upper limit of the expected one-time investment cost input at the nth stage;

[0029] The unit commitment constraints include wind-solar operation constraints, thermal power operation constraints, electrochemical energy storage equipment operation constraints, and spinning reserve capacity constraints;

[0030] The wind-solar operation constraints are:

[0031]

[0032] wherein, is the actual output of the wind power at the nth stage, s-th scenario, and t-th time period; is the actual output of the photovoltaic power at the nth stage, s-th scenario, and t-th time period; is the installed capacity of the wind power at the nth stage; is the installed capacity of the photovoltaic power at the nth stage; is the predicted capacity factor of the wind power at the nth stage, s-th scenario, and t-th time period; is the predicted capacity factor for the nth stage, s-th scenario, and t-th time period of photovoltaic;

[0033] The thermal power operation constraint is:

[0034]

[0035] where, is the actual output of thermal power for the nth stage, s-th scenario, and t-th time period; is the actual output of thermal power for the nth stage, s-th scenario, and (t - 1)-th time period; is the ramp power coefficient of thermal power; is the grid-connected capacity of thermal power for the nth stage, s-th scenario, and t-th time period, is the down-ramp power coefficient of thermal power; is the minimum technical output coefficient of thermal power; is the maximum technical output coefficient of thermal power; is the total installed capacity of thermal power, is the installed capacity of the j-th unit for the nth stage of thermal power; is the start-stop state of the j-th unit for the nth stage, s-th scenario, and t-th time period of thermal power;

[0036] The operation constraint of the electrochemical energy storage device is:

[0037]

[0038] where, is the external equivalent actual output of the electrochemical energy storage device for the nth stage, s-th scenario, and t-th time period; is the discharge power of the electrochemical energy storage device for the nth stage, s-th scenario, and t-th time period; is the charging power of the electrochemical energy storage device for the nth stage, s-th scenario, and t-th time period; is the installed capacity of the electrochemical energy storage device for the nth stage; is the state of charge of the electrochemical energy storage device for the nth stage, s-th scenario, and t-th time period; is the state of charge of the electrochemical energy storage device for the nth stage, s-th scenario, and (t - 1)-th time period; is the charging efficiency of the electrochemical energy storage device for the nth stage; is the discharge efficiency of the electrochemical energy storage device for the nth stage; is the rated energy of the electrochemical energy storage device for the nth stage; Δt is the time period interval; is the minimum state of charge; is the maximum state of charge; is the initial state of charge for the nth stage, s-th scenario; is the state of charge at the last time period for the nth stage, s-th scenario;

[0039] The rotational reserve capacity constraint is as follows:

[0040]

[0041] Wherein, is the output power of the electrochemical energy storage device at the t-th time period of the s-th scenario in the n-th stage; is the positive reserve coefficient of wind power; is the positive reserve coefficient of photovoltaic power; is the positive reserve coefficient of electrical load; is the power of the electrical load at the t-th time period of the s-th scenario in the n-th stage; is the negative reserve coefficient of wind power; is the negative reserve coefficient of photovoltaic power; is the negative reserve coefficient of electrical load;

[0042] The clean and low-carbon driving constraints include carbon emission limit constraints, new energy electricity proportion constraints, and minimum new energy consumption rate constraints;

[0043]

[0044] Wherein, τ s is the weight coefficient of the s-th scenario; is the carbon emission factor of thermal power in the n-th stage; is the carbon emission limit in the n-th stage, is the new energy electricity proportion requirement in the n-th stage; is the expected new energy consumption rate in the n-th stage.

[0045] Optionally, based on the initial installed capacity domain, using the wave search algorithm, determine the installed capacity domain labels corresponding to the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate for each stage of the training power system, specifically including:

[0046] Normalize the maximum total cost and the maximum new energy consumption rate corresponding to the optimal solution set of the installed capacity domain to obtain the total cost reference value and the new energy consumption rate reference value;

[0047] Solve the reduced process objective function according to the total cost reference value and the new energy consumption rate reference value to obtain the reduced process reference objective value; the reduced process objective function is Loss = δ f f a + δ g g a ; wherein, Loss is the reduced process reference objective value, f a is the total cost reference value; g a is the new energy consumption rate reference value; δ f is the economic index weight; δg is the weight of the accommodation rate index;

[0048] According to the total cost benchmark value, the new energy accommodation rate benchmark value, and the reduction process benchmark target value, using the wave search algorithm, determine the upper limit of the one-time investment coefficient input and the installed capacity domain label corresponding to the lower limit of the new energy accommodation rate for each stage of the power system for training.

[0049] In a second aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the power system installed capacity domain generation method described in any one of the above.

[0050] In a third aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the power system installed capacity domain generation method described in any one of the above.

[0051] In a fourth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the power system installed capacity domain generation method described in any one of the above.

[0052] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0053] The present application provides a method, device, medium, and product for generating the installed capacity domain of a power system, obtains the expected value of the upper limit of the one-time investment coefficient input and the expected value of the lower limit of the new energy accommodation rate of the target power system, and uses the power system installed capacity domain generation model to determine the installed capacity domain of the target power system; the installed capacity domain is composed of the upper bound of the installed capacity domain and the lower bound of the installed capacity domain; among them, the power system installed capacity domain generation model is obtained by training the CNN-BiLSTM-BiGRU neural network model using a training data set; the training data set is established using a multi-objective coronavirus disease optimization algorithm and a wave search algorithm; the CNN-BiLSTM-BiGRU neural network model includes a convolutional neural network, a bidirectional long short-term memory neural network, and a bidirectional gated recurrent unit connected in sequence. The present application improves the accuracy of generating the installed capacity domain of the power system. Description of the Drawings

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0055] Figure 1 Schematic flow chart of a method for generating the installed capacity domain of a power system provided by an embodiment of the present application;

[0056] Figure 2 Architecture diagram of the method for generating the installed capacity domain of the power system of the present application;

[0057] Figure 3 Overall flow chart of the method for generating the installed capacity domain of the power system of the present application;

[0058] Figure 4 Schematic structural diagram of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0059] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0060] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0061] In an exemplary embodiment, as Figure 1 and Figure 2 shown, a method for generating the installed capacity domain of a power system is provided, including the following steps:

[0062] S1: Obtain the upper limit expected value of the one-time investment coefficient input and the lower limit expected value of the new energy consumption rate of the target power system.

[0063] S2: According to the upper limit expected value of the one-time investment coefficient input and the lower limit expected value of the new energy consumption rate, use the installed capacity domain generation model of the power system to determine the installed capacity domain of the target power system; the installed capacity domain is composed of an upper bound of the installed capacity domain and a lower bound of the installed capacity domain; wherein, the installed capacity domain generation model of the power system is obtained by training the CNN-BiLSTM-BiGRU neural network model using a training data set; the training data set is established using a multi-objective coronavirus disease optimization algorithm and a wave search algorithm; the training data set includes the upper limit of the one-time investment coefficient input, the lower limit of the new energy consumption rate, and the corresponding installed capacity domain labels at each stage of the training power system; the CNN-BiLSTM-BiGRU neural network model includes a convolutional neural network, a bidirectional long short-term memory neural network, and a bidirectional gated recurrent unit connected in sequence.

[0064] The first layer of the CNN-BiLSTM-BiGRU neural network model uses a Convolutional Neural Network (CNN) and performs a max-pooling transformation. Among them, each updated feature vector is considered a new input feature. The convolutional layer and the max-pooling layer achieve the aggregation of feature vectors and output the maximum value, effectively reducing the data dimension; the second layer uses BiLSTM (Bidirectional Long Short-Term Memory) to capture the long-term bidirectional interaction between the time steps of the sequence data. A Dropout layer is added after the BiLSTM layer to reduce overfitting; the third layer uses BiGRU (Bidirectional Gated Recurrent Unit), and BiGRU has a sequence processing model with two GRUs that can process data forward and backward.

[0065] As an alternative implementation, the CNN-BiLSTM-BiGRU neural network model is trained using a training data set, specifically including:

[0066] S21: Construct a training data set.

[0067] This application also proposes a power system installed capacity domain based on domain theory.

[0068] First of all, domain theory uses projection observation technology to present the implicit mathematical analysis expression of the system operation model in spatial geometry according to different analysis perspectives, which is an important method to support the construction of the power system operation space. Using domain theory, starting from the perspective of the installed capacity of wind, light, fire, and storage (wind power, photovoltaic power, thermal power, and electrochemical energy storage) in the power system, a power system installed capacity domain is established, which is defined as: considering the power load growth demand and the characteristics of the power source structure of the power system, considering the system economy and the improvement of the new energy consumption capacity, the set of decision points that meet the power balance, reliable operation, and clean and low-carbon driving constraints on a long-term scale, where the decision point is the annual installed capacity of various types of power source equipment. The expression of the power system installed capacity domain model is:

[0069]

[0070] Where: Ω is the installed capacity domain of the power system; x is the key decision vector corresponding to the system decision point within the long-term scale, including the installed capacity of various types of power generation equipment at each stage; z represents the discrete variables of the system, including the start-stop status of thermal power units and the charge-discharge status of electrochemical energy storage devices at each time period, etc.; y represents the vector composed of the remaining continuous variables of the system, including the specific output status of each unit and other auxiliary variables. f(x) ≤ 0 is the unit capacity constraint; g(x, y, z) ≤ 0 represents the unit commitment constraint; h(y, z) = 0 is the power balance constraint; k(y, z) ≤ 0 is the clean and low-carbon driving constraint.

[0071] Secondly, for the above-mentioned various types of constraints, the specific constraint expressions of the installed capacity domain model of the power system are as follows.

[0072] 1) Constraint type 1: The unit capacity constraint mainly includes the upper and lower limits of the unit capacity, the connection constraint of the unit capacity, and the upper limit of the one-time investment cost input.

[0073]

[0074] Where: is the installed capacity of the power generation equipment in the nth stage; is the installed capacity of the power generation equipment in the (n + 1)th stage; is the lower limit of the installed capacity in the nth stage; is the upper limit of the installed capacity in the nth stage; is the unit capacity investment cost of the power generation equipment in the nth stage, is the installed capacity of the power generation equipment in the (n - 1)th stage; is the upper limit of the expected one-time investment cost input in the nth stage.

[0075] 2) Constraint type 2: The unit commitment constraint mainly includes the operation constraints of wind-solar-thermal-energy storage equipment and the spinning reserve capacity constraint.

[0076] Wind-solar operation constraint:

[0077]

[0078] Where: is the actual output of wind power in the nth stage, the sth scenario, and the tth time period; is the actual output of photovoltaic power in the nth stage, the sth scenario, and the tth time period; is the installed capacity of wind power in the nth stage; is the installed capacity of photovoltaic power in the nth stage; is the predicted capacity factor of wind power in the nth stage, the sth scenario, and the tth time period; is the predicted capacity factor of photovoltaic power in the nth stage, the sth scenario, and the tth time period.

[0079] Thermal power operation constraint:

[0080]

[0081] Wherein: is the actual output of thermal power at the nth stage, s-th scenario, and t-th time period; is the actual output of thermal power at the nth stage, s-th scenario, and (t - 1)-th time period; is the ramp power coefficient of thermal power; is the grid connection capacity of thermal power at the nth stage, s-th scenario, and t-th time period, is the down-ramp power coefficient of thermal power; is the minimum technical output coefficient of thermal power; is the maximum technical output coefficient of thermal power; is the total installed capacity of thermal power, is the installed capacity of the j-th unit at the nth stage of thermal power; is the start-stop state of the j-th unit at the nth stage, s-th scenario, and t-th time period of thermal power.

[0082] Electrochemical energy storage device operation constraint:

[0083]

[0084] Further linearize using the big M method, introducing binary auxiliary variables

[0085]

[0086] Wherein: is the external equivalent actual output of the electrochemical energy storage device at the nth stage, s-th scenario, and t-th time period; is the discharge power of the electrochemical energy storage device at the nth stage, s-th scenario, and t-th time period; is the charging power of the electrochemical energy storage device at the nth stage, s-th scenario, and t-th time period; is the installed capacity of the electrochemical energy storage device at the nth stage; is the state of charge of the electrochemical energy storage device at the nth stage, s-th scenario, and t-th time period; is the state of charge of the electrochemical energy storage device at the nth stage, s-th scenario, and (t - 1)-th time period; is the charging efficiency of the electrochemical energy storage device at the nth stage; is the discharge efficiency of the electrochemical energy storage device at the nth stage; is the rated energy of the electrochemical energy storage device at the nth stage; Δt is the time interval; is the minimum state of charge; is the maximum state of charge; is the initial state of charge at the nth stage, s-th scenario; is the state of charge at the end of the last time period in the s-th scenario of the n-th stage; is the charge and discharge status of the electrochemical energy storage device in the t-th time period of the s-th scenario of the n-th stage; M is a constant coefficient.

[0087] Reserve capacity constraint:

[0088]

[0089] Where: is the output power of the electrochemical energy storage device in the t-th time period of the s-th scenario of the n-th stage; is the positive reserve coefficient of wind power; is the positive reserve coefficient of photovoltaic power; is the positive reserve coefficient of electrical load; is the power of the electrical load in the t-th time period of the s-th scenario of the n-th stage; is the negative reserve coefficient of wind power; is the negative reserve coefficient of photovoltaic power; is the negative reserve coefficient of electrical load.

[0090] 3) Constraint type 3: Power balance constraint

[0091]

[0092] Where: is the actual electrical load.

[0093] 4) Constraint type 4: Clean and low-carbon driving constraints mainly include carbon emission limit constraints, new energy electricity proportion constraints, and minimum new energy consumption rate constraints.

[0094]

[0095] Where: τ s is the weight coefficient of the s-th scenario; is the carbon emission factor of thermal power in the n-th stage; is the carbon emission limit in the n-th stage, is the new energy electricity proportion requirement in the n-th stage; is the expected new energy consumption rate in the n-th stage.

[0096] In practical applications, the specific process of establishing the installed capacity domain dataset (training dataset) is as follows: Generate the upper limit of the one-time investment cost and the lower limit of the new energy consumption rate for each stage through a random function. Further, adopt the installed capacity domain generation method based on the multi-objective coronavirus disease optimization algorithm and the installed capacity domain boundary improvement method based on the wave search algorithm to obtain the corresponding upper and lower bound data of the installed capacity domain. The massive input data and output data generated are used to form the installed capacity domain dataset. A set of data in the installed capacity domain dataset includes the upper limit of the one-time investment cost for each stage, the lower limit of the new energy consumption rate, the upper bound of the installed capacity domain, and the lower bound of the installed capacity domain.

[0097] As an alternative implementation method, constructing the training dataset specifically includes:

[0098] S211: Use a random function to generate the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate for each stage of the power system for training.

[0099] S212: Use the multi-objective coronavirus disease optimization algorithm to generate the initial installed capacity domain of the power system for training, specifically including:

[0100] Step 1: Establish a multi-objective optimization model; the multi-objective optimization model takes the minimum total coefficient and the minimum new energy consumption rate as the objective functions, and takes the unit capacity constraint of the unit, the unit combination constraint, and the clean and low-carbon driving constraint as the constraint conditions.

[0101] The new energy consumption rate is an important indicator for the assessment of the current power system, and a reasonable power source structure arrangement on the source side is the key factor affecting the new energy consumption target. To enhance the potential of the installed capacity domain in terms of new energy consumption, a multi-objective optimization model considering the new energy consumption rate is established, and the installed capacity domain is generated using the Pareto solution set of the multi-objective problem.

[0102] The multi-objective optimization model considering the new energy consumption rate is as follows:

[0103]

[0104] Among them, F a is the total cost; T a is the number of days in the current stage; is the unit capacity maintenance cost of the power supply equipment in the nth stage; is the unit power operation cost of the electrochemical energy storage equipment in the nth stage; is the unit power operation cost of thermal power in the nth stage; is the unit power penalty cost of load shedding, is the actual load shedding at the s-th scenario and t-th time period in the nth stage; G a is the new energy consumption rate, N rThe number of stages for power capacity decision-making.

[0105] Step 2: Use the multi-objective coronavirus disease optimization algorithm to solve the multi-objective optimization model and determine the optimal solution set of the installed capacity domain.

[0106] Secondly, for the above multi-objective optimization problem, a multi-objective coronavirus disease optimization algorithm is used for rapid solution. This algorithm is based on the coronavirus disease optimization algorithm. During the optimization process, an archive is used to store non-dominated Pareto optimal solutions. By simulating the process of coronavirus replication, a roulette wheel selection strategy is adopted to select effective archive solutions, and it has excellent effects in solving the global optimization problem of two objective functions.

[0107] The coronavirus disease optimization algorithm is a heuristic optimization algorithm inspired by the replication mechanism of coronaviruses hijacking human cells. It mainly consists of four stages: virus entry and stripping, virus replication, virus mutation, and new virus release. The search mechanism of the multi-objective coronavirus disease optimization algorithm is consistent with that of the coronavirus disease optimization algorithm. In particular, the multi-objective coronavirus disease optimization algorithm uses a dominance operator to compare solutions considering multiple objective functions. All Pareto optimal solutions obtained during the optimization process are stored in an archive, and an archive controller is used to decide whether to retain or delete the solutions in the archive.

[0108] The principle of the archive controller is as follows:

[0109] 1) If the archive is empty, the current solution should be accepted.

[0110] 2) If another solution dominates in the archive, the corresponding solution in the archive should be deleted.

[0111] 3) If another solution does not dominate, the corresponding solution should be stored in the archive.

[0112] To improve the coverage of Pareto optimal solutions, solutions must be selected from the least crowded regions of the archive to promote the improvement of other regions, and many neighboring solutions should be deleted from the archive. MOVOIDOA uses a roulette wheel selection strategy to select a solution from the region with the fewest solutions in the archive; when the number of solutions in the archive reaches the upper limit, the solution with many neighboring solutions should be deleted. The probabilities of the two selections are as follows:

[0113]

[0114] Where: P i O is the probability of selecting a new solution, P i D is the probability of deleting a solution; c MO is a constant, N iMO Denotes the number of neighboring solutions of the \(i\)-th solution.

[0115] Finally, the multi-objective coronavirus disease optimization algorithm solves the multi-objective optimization model considering the new energy consumption rate, and obtains the Pareto optimal solution set of the multi-objective optimization problem considering the new energy consumption rate. Among all solutions, the maximum power supply capacity is selected as the upper bound, and the minimum power supply capacity is selected as the lower bound to obtain the initial installed capacity domain of each type of power supply that expands with the increase of each stage.

[0116] Step 3: Determine the initial installed capacity domain of the training power system according to the optimal solution set of the installed capacity domain; the upper bound of the initial installed capacity domain is the maximum value of the power supply capacity in the optimal solution set of the installed capacity domain; the lower bound of the initial installed capacity domain is the minimum value of the power supply capacity in the optimal solution set of the installed capacity domain.

[0117] S213: Based on the initial installed capacity domain, use the wave search algorithm to determine the installed capacity domain label corresponding to the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate for each stage of the training power system, and obtain the training data set.

[0118] In practical applications, a method for improving the boundary of the installed capacity domain based on the wave search algorithm is proposed. By using the virtual boundary and iterative optimization method, the economy and the new energy consumption rate are integrated, and the installed capacity domain is gradually reduced to effectively improve the effect of the installed capacity domain.

[0119] As an optional implementation manner, based on the initial installed capacity domain, use the wave search algorithm to determine the installed capacity domain label corresponding to the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate for each stage of the training power system, specifically including:

[0120] (1) Normalize the maximum total cost and the maximum new energy consumption rate corresponding to the optimal solution set of the installed capacity domain to obtain the total cost reference value and the new energy consumption rate reference value.

[0121] First, extract the maximum total cost and the maximum new energy consumption rate in the Pareto optimal solution set (optimal solution set of the installed capacity domain) in step 2, and the economic index (corresponding total cost) and the consumption rate index (corresponding new energy consumption rate) of the capacity design scheme can be normalized.

[0122] Furthermore, use the fairness principle to extract a set of capacity design schemes in the Pareto optimal solution set, and regard the total cost and the new energy consumption rate under this scheme as the reference values. After normalization, the reference economic index (total cost reference value) and the reference consumption rate index (new energy consumption rate reference value) are obtained.

[0123]

[0124] Where: ΔF a is the total cost increment; is the minimum value of the total cost; is the maximum value of the total cost, ΔG a is the increment of the new energy consumption rate; is the minimum value of the increment of the new energy consumption rate; is the maximum value of the increment of the new energy consumption rate; is the normalized benchmark value of the total cost; is the normalized benchmark value of the new energy consumption rate.

[0125] (2) Solve the reduced process objective function according to the total cost benchmark value and the new energy consumption rate benchmark value to obtain the reduced process benchmark target value; the reduced process objective function is Loss = δ f f a + δ g g a ; where Loss is the reduced process benchmark target value, f a is the total cost benchmark value; g a is the new energy consumption rate benchmark value; δ f is the economic index weight; δ g is the consumption rate index weight.

[0126] Secondly, use the method of virtual boundary and iterative optimization to propose a reduction process for the installed capacity domain.

[0127] The objective function of the installed capacity domain reduction process is as follows, and the reduced process benchmark target value can be obtained according to the benchmark economic index and the benchmark consumption rate index.

[0128] Loss = δ f f a + δ g g a (14)

[0129] Where: f a is the total cost benchmark value; g a is the new energy consumption rate benchmark value; δ f is the economic index weight; δ g is the consumption rate index weight.

[0130] (3) According to the total cost benchmark value, the new energy consumption rate benchmark value and the reduced process benchmark target value, use the wave search algorithm to determine the installed capacity domain label corresponding to the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate for each stage of the training power system.

[0131] 1) The capacities of various types of power supply equipment are initialized in the initial installed capacity domain using a random function, and the reduction process objective value is calculated under the reduction process objective function and compared with the reduction process benchmark objective value.

[0132] 2) The upper and lower bounds of the power installed capacity domain are described using virtual boundaries.

[0133] If the reduction process objective value is worse than the reduction process benchmark objective value, the virtual boundary sub-elements are corrected according to the distances from the upper and lower bounds of the initial installed capacity domain. The virtual boundary sub-elements select the average with the adjacent initial installed capacity domain boundary elements to generate new virtual upper and lower bounds.

[0134] If the reduction process objective value is better than the reduction process benchmark objective value, the virtual boundary sub-elements start from the current capacity, and the virtual boundary repels to both sides to generate new virtual upper and lower bounds, and the repulsion range coefficient is determined according to the excellence degree of the current objective value.

[0135]

[0136] Where: R ex is the repulsion range coefficient, α ex is a constant coefficient, Loss bo is the reduction process benchmark objective value.

[0137] 3) The wave search algorithm is used to continuously update the power supply equipment capacity and generate new virtual boundaries until the fluctuation range of the virtual boundary is less than the set threshold, and the iteration ends.

[0138]

[0139] Where: N r is the number of stages of the power capacity decision, γ k is the set threshold coefficient.

[0140] Finally, the wave search algorithm is used to determine the iteration direction to accelerate convergence.

[0141] The WSA (Wave Search Algorithm) utilizes a unique algorithm design concept of radar technology, adopts various improved greedy mechanisms, and utilizes the gradient information of the problem to be optimized, making the algorithm characterized by high precision, high efficiency, and strong adaptability. This application uses the wave search algorithm to determine the iteration direction, which is specifically divided into three stages: emitting electromagnetic waves, reflecting electromagnetic waves, and receiving electromagnetic waves. Further, a fitting gradient descent method based on the central difference method is introduced to improve the search efficiency and accuracy. The specific process is to update the positions in the three stages of emitting electromagnetic waves, reflecting electromagnetic waves, and receiving electromagnetic waves in sequence according to the initialized positions of electromagnetic wave particles, and further improve the position information through the fitting gradient descent method based on the central difference method to complete this iteration; continuously update the positions iteratively until the upper limit of the iteration times is reached.

[0142] Strategy for the stage of emitting electromagnetic waves:

[0143]

[0144] where: σ w is the waveform control coefficient, is the i-th element of the column vector that follows the normal distribution and is arranged in order, X best is the current optimal position, X i is the current position, is the position matrix rearranged according to the degree of proximity to X best is the position matrix rearranged according to the degree of proximity to X is the newly generated position, L max is the maximum objective value within the population.

[0145] The function of formula (17) is to simulate the outward diffusion of electromagnetic waves, reduce the possibility of falling into local optimality, and improve the search efficiency. Formula (18) is an improved greedy mechanism, and its function is to make the population position not inferior to the current population position when the population position fluctuates outward.

[0146] Strategy for the stage of reflecting electromagnetic waves:

[0147]

[0148] where: β w is the reflection intensity coefficient, r1 w is a random value between 0 and 1, is the number of particles simulating the reflected electromagnetic waves. is the position matrix rearranged in ascending order of the objective value.

[0149] Strategy for the stage of receiving electromagnetic waves: The judgment strategy for updating the electromagnetic wave particles in the stage of receiving electromagnetic waves is the same as formula (20) in the stage of reflecting electromagnetic waves.

[0150]

[0151] where: δ w is the reception coefficient, η w is a random number following a normal distribution, is the number of particles for simulating the received electromagnetic wave. is the uniform optimal position, λ w is the correction factor, is a random value between 0 and 1.

[0152] Furthermore, a deterministic optimization technique is introduced, namely the fitting gradient descent method based on the central difference method. This method uses the central difference method to fit the analytical information of the problem to be optimized, and is used to search for the optimal solution to improve the search efficiency and accuracy. Its mathematical expression is:

[0153]

[0154] where: ε = 10 -6 , is the i-th element of the gradient, α w is the step size coefficient.

[0155] S22: Input the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate at each stage of the training power system into the current CNN-BiLSTM-BiGRU neural network model to obtain the upper bound prediction value and the lower bound prediction value of the installed capacity domain.

[0156] S23: Determine the loss function value according to the upper bound prediction value of the installed capacity domain, the lower bound prediction value of the installed capacity domain, and the corresponding installed capacity domain label.

[0157] S24: Determine whether the end training condition is satisfied; the end training condition is to reach the maximum number of iterations or the loss function value is less than the preset value.

[0158] S25: If so, end the training and use the current CNN-BiLSTM-BiGRU neural network model as the installed capacity domain generation model for the power system.

[0159] S26: If not, adjust the model parameters of the current CNN-BiLSTM-BiGRU neural network model according to the loss function value, and return "Input the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate at each stage of the training power system into the current CNN-BiLSTM-BiGRU neural network model to obtain the upper bound prediction value and the lower bound prediction value of the installed capacity domain".

[0160] In practical applications, the installed capacity domain dataset is input into the CNN-BiLSTM-BiGRU neural network model for continuous training. After the training is completed, a finished neural network model is obtained. By inputting the upper limit of the expected one-time investment cost and the lower limit of the new energy consumption rate into the CNN-BiLSTM-BiGRU neural network model, the upper bound and lower bound of the installed capacity domain can be quickly obtained.

[0161] A method for generating the installed capacity domain of a power system proposed in this application, as Figure 3 shown. First, the installed capacity domain of the power system is proposed based on domain theory. Second, a multi-objective optimization model considering the new energy consumption capacity is established, and the multi-objective coronavirus disease optimization algorithm is used to solve it to generate the initial installed capacity domain. Third, a process for reducing the installed capacity domain is proposed using virtual boundaries and iterative optimization, and the wave search algorithm is used to determine the iterative direction to generate the reduced installed capacity domain. Finally, based on the installed capacity domain dataset, a CNN-BiLSTM-BiGRU neural network model is constructed to achieve the rapid generation of the installed capacity domain.

[0162] The method for generating the installed capacity domain of a power system proposed in this application has the following advantages:

[0163] (1) The installed capacity domain of the power system proposed in this application based on domain theory can comprehensively describe the multi-dimensional attributes and constraint conditions of the power system capacity, clearly display the boundaries and potential bottlenecks of the power system capacity, and provide an intuitive and powerful decision support tool for decision-makers.

[0164] (2) This application uses the multi-objective coronavirus disease optimization algorithm to achieve the rapid solution of the multi-objective optimization problem considering the new energy consumption capacity. Using the excellent global search and local optimization capabilities of the algorithm, it can achieve rapid optimization under complex multi-objective problems, thereby improving the efficiency and effect of optimization.

[0165] (3) This application proposes a process for reducing the installed capacity domain using virtual boundaries and iterative optimization, and uses the wave search algorithm to determine the iterative direction. By using the dynamic characteristics of waves for search and optimization, it can quickly cover the search space, accelerate the convergence process, and improve the calculation efficiency.

[0166] (4) This application uses the CNN-BiLSTM-BiGRU neural network model, which combines the advantages of CNN, BiLSTM, and BiGRU, has strong time series feature extraction and processing capabilities, can comprehensively capture the complex relationships in the data. The finished neural network model can quickly generate the installed capacity domain boundary under the input data, and significantly improve the accuracy of generating the installed capacity domain, and is expected to provide strong decision support for the efficient management and optimization of the power system power structure.

[0167] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structural diagram can be as shown in Figure 4 Shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a method for generating a power system installed capacity domain.

[0168] Those skilled in the art can understand that Figure 4 The structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0169] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the above-mentioned method for generating a power system installed capacity domain is implemented.

[0170] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by the processor, the above-mentioned method for generating a power system installed capacity domain is implemented.

[0171] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by the processor, the above-mentioned method for generating a power system installed capacity domain is implemented.

[0172] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0173] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the various embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAMs), magnetoresistive random access memories (MRAMs), ferroelectric random access memories (FRAMs), phase change memories (PCMs), graphene memories, etc. Volatile memories can include random access memories (RAMs) or external caches, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0174] The databases involved in the various embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the various embodiments provided in this application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0175] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0176] Specific examples are used in this article to elaborate on the principles and implementation manners of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. A method for generating the installed capacity domain of a power system, characterized in that Including: Obtaining the expected upper limit of the one-time investment coefficient input and the expected lower limit of the new energy consumption rate of the target power system; According to the expected upper limit of the one-time investment coefficient input and the expected lower limit of the new energy consumption rate, using the power system installed capacity domain generation model to determine the installed capacity domain of the target power system; the installed capacity domain is composed of an installed capacity domain upper bound and an installed capacity domain lower bound; wherein, the power system installed capacity domain generation model is obtained by training a CNN-BiLSTM-BiGRU neural network model using a training data set; the training data set is established using a multi-objective coronavirus disease optimization algorithm and a wave search algorithm; the training data set includes the upper limit of the one-time investment coefficient input, the lower limit of the new energy consumption rate at each stage of the training power system, and the corresponding installed capacity domain labels; the CNN-BiLSTM-BiGRU neural network model includes a convolutional neural network, a bidirectional long short-term memory neural network, and a bidirectional gated recurrent unit connected in sequence; Constructing a training data set, specifically including: Using a random function to generate the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate at each stage of the training power system; Using a multi-objective coronavirus disease optimization algorithm to generate the initial installed capacity domain of the training power system; Based on the initial installed capacity domain, using a wave search algorithm to determine the installed capacity domain labels corresponding to the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate at each stage of the training power system, obtaining a training data set; Using a multi-objective coronavirus disease optimization algorithm to generate the initial installed capacity domain of the training power system, specifically including: Establishing a multi-objective optimization model; the multi-objective optimization model takes the minimum total coefficient and the minimum new energy consumption rate as objective functions, and takes unit capacity constraints, unit combination constraints, and clean and low-carbon drive constraints as constraint conditions; Using a multi-objective coronavirus disease optimization algorithm to solve the multi-objective optimization model to determine the optimal solution set of the installed capacity domain; According to the optimal solution set of the installed capacity domain, determining the initial installed capacity domain of the training power system; the upper bound of the initial installed capacity domain is the maximum value of the power supply capacity in the optimal solution set of the installed capacity domain; the lower bound of the initial installed capacity domain is the minimum value of the power supply capacity in the optimal solution set of the installed capacity domain.

2. The method for generating the installed capacity domain of the power system according to claim 1, wherein, Using a training data set to train a CNN-BiLSTM-BiGRU neural network model, specifically including: Constructing a training data set; Inputting the upper limit of the one-time investment coefficient input and the lower limit of the new energy consumption rate at each stage of the training power system into the current CNN-BiLSTM-BiGRU neural network model to obtain a predicted value of the installed capacity domain upper bound and a predicted value of the installed capacity domain lower bound; Determining the loss function value according to the predicted value of the installed capacity domain upper bound, the predicted value of the installed capacity domain lower bound, and the corresponding installed capacity domain label; Judging whether the end training condition is satisfied; the end training condition is reaching the maximum number of iterations or the loss function value being less than a preset value; If so, end the training and use the current CNN-BiLSTM-BiGRU neural network model as the power system installed capacity domain generation model; If not, adjust the model parameters of the current CNN-BiLSTM-BiGRU neural network model according to the loss function value, and return "Input the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate at each stage of the training power system into the current CNN-BiLSTM-BiGRU neural network model to obtain the predicted upper bound value and the predicted lower bound value of the installed capacity domain".

3. The method for generating the installed capacity domain of the power system according to claim 1, characterized in that The objective function is: Among them, F a is the total cost; T a is the number of days in the current stage; is the unit capacity investment cost of the power supply equipment in the nth stage; is the installed capacity of the power supply equipment in the nth stage; is the installed capacity of the power supply equipment in the (n + 1)th stage; is the unit capacity maintenance cost of the power supply equipment in the nth stage; τ s is the weight coefficient of the sth scenario; is the unit power operation cost of the electrochemical energy storage equipment in the nth stage; is the charging power of the electrochemical energy storage equipment in the nth stage, sth scenario, and tth time period; is the discharging power of the electrochemical energy storage equipment in the nth stage, sth scenario, and tth time period; Δt is the time period interval; is the unit power operation cost of thermal power in the nth stage; is the actual output of thermal power in the nth stage, sth scenario, and tth time period; is the unit power penalty cost of load shedding, is the actual load shedding in the nth stage, sth scenario, and tth time period; G a is the new energy consumption rate, N r is the number of stages of power capacity decision-making; is the actual output of wind power in the nth stage, sth scenario, and tth time period; is the actual output of photovoltaic power in the nth stage, sth scenario, and tth time period; is the installed capacity of wind power in the nth stage; is the installed capacity of photovoltaic power in the nth stage; is the predicted capacity factor of wind power in the nth stage, sth scenario, and tth time period; is the predicted capacity factor of photovoltaic power in the nth stage, sth scenario, and tth time period.

4. The method for generating the installed capacity domain of the power system according to claim 1, wherein The unit capacity constraint is: Among them, is the installed capacity of power equipment in the nth stage; is the installed capacity of power equipment in the (n + 1)th stage; is the lower limit of the installed capacity in the nth stage; is the upper limit of the installed capacity in the nth stage; is the investment cost per unit capacity of the power equipment in the nth stage, is the installed capacity of power equipment in the (n - 1)th stage; is the upper limit of the expected one-time investment cost input in the nth stage; The unit combination constraints include wind and light operation constraints, thermal power operation constraints, electrochemical energy storage device operation constraints, and spinning reserve capacity constraints; The wind and light operation constraints are: Among them, is the actual output of wind power in the t-th time period of the s-th scenario in the n-th stage; is the actual output of photovoltaic power in the t-th time period of the s-th scenario in the n-th stage; is the installed capacity of wind power in the n-th stage; is the installed capacity of photovoltaic power in the n-th stage; is the predicted capacity factor of wind power in the t-th time period of the s-th scenario in the n-th stage; is the predicted capacity factor of photovoltaic power in the t-th time period of the s-th scenario in the n-th stage; The thermal power operation constraints are: Among them, is the actual output of thermal power at the nth stage, s-th scenario, and t-th time period; is the actual output of thermal power at the nth stage, s-th scenario, and (t - 1)-th time period; is the ramp power coefficient of thermal power; is the grid-connected capacity of thermal power at the nth stage, s-th scenario, and t-th time period, is the down-ramp power coefficient of thermal power; is the minimum technical output coefficient of thermal power; is the maximum technical output coefficient of thermal power; is the total installed capacity of thermal power, is the installed capacity of the j-th unit at the nth stage of thermal power; is the start-stop state of the j-th unit at the nth stage, s-th scenario, and t-th time period of thermal power; The electrochemical energy storage device operation constraints are: Among them, is the external equivalent actual output of the electrochemical energy storage device in the t-th time period of the s-th scenario in the n-th stage; is the discharge power of the electrochemical energy storage device in the t-th time period of the s-th scenario in the n-th stage; is the charging power of the electrochemical energy storage device in the t-th time period of the s-th scenario in the n-th stage; is the installed capacity of the electrochemical energy storage device in the n-th stage; is the state of charge of the electrochemical energy storage device in the t-th time period of the s-th scenario in the n-th stage; is the state of charge of the electrochemical energy storage device in the (t - 1)-th time period of the s-th scenario in the n-th stage; is the charging efficiency of the electrochemical energy storage device in the n-th stage; is the discharge efficiency of the electrochemical energy storage device in the n-th stage; is the rated energy of the electrochemical energy storage device in the n-th stage; Δt is the time interval; is the minimum state of charge; is the maximum state of charge; is the initial state of charge of the s-th scenario in the n-th stage; is the state of charge at the last time period of the s-th scenario in the n-th stage; The spinning reserve capacity constraints are: Among them, is the output power of the electrochemical energy storage device at the t-th time period of the s-th scenario in the n-th stage; is the positive reserve coefficient of wind power; is the positive reserve coefficient of photovoltaic power; is the positive reserve coefficient of the electrical load; is the power of the electrical load at the t-th time period of the s-th scenario in the n-th stage; is the negative reserve coefficient of wind power; is the negative reserve coefficient of photovoltaic power; is the negative reserve coefficient of the electrical load; The clean and low-carbon driving constraints include carbon emission limit constraints, new energy electricity proportion constraints, and minimum new energy consumption rate constraints; Among them, τ s is the weight coefficient of the sth scenario; is the carbon emission factor of the nth stage of thermal power; is the carbon emission limit of the nth stage, is the requirement for the proportion of new energy electricity in the nth stage; is the expected new energy consumption rate in the nth stage.

5. The method for generating the installed capacity domain of the power system according to claim 1, wherein Based on the initial installed capacity domain, use the wave search algorithm to determine the installed capacity domain labels corresponding to the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate at each stage of the training power system, specifically including: Normalize the maximum total cost and the maximum new energy consumption rate corresponding to the optimal solution set of the installed capacity domain to obtain the total cost reference value and the new energy consumption rate reference value; Solve the reduction process objective function according to the total cost benchmark value and the new energy consumption rate benchmark value to obtain the reduction process benchmark objective value; the reduction process objective function is Loss = δ f f a + δ g g a ; where Loss is the reduction process benchmark objective value, f a is the total cost benchmark value; g a is the new energy consumption rate benchmark value; δ f is the economic index weight; δ g is the consumption rate index weight; According to the total cost reference value, the new energy consumption rate reference value, and the reduction process reference target value, use the wave search algorithm to determine the installed capacity domain labels corresponding to the upper limit of the one-time investment coefficient and the lower limit of the new energy consumption rate at each stage of the training power system.

6. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the power system installed capacity domain generation method according to any one of claims 1-5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the power system installed capacity domain generation method according to any one of claims 1-5.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the power system installed capacity domain generation method according to any one of claims 1-5.

Images