Time-series production simulation method taking capacity dynamic programming into consideration, and apparatus and medium

Abstract

A time-series production simulation method taking capacity dynamic programming into consideration, and an apparatus and a medium, relating to the technical field of power systems. The method comprises: establishing a single-year time-series production simulation model; on the basis of the single-year time-series production simulation model and a CNN-GRU network, constructing an equivalent load support rate interval decision network; on the basis of the single-year time-series production simulation model, constructing a multi-stage dynamic programming model oriented to the problem of power capacity expansion; using a NO search algorithm to compress a state space in the multi-stage dynamic programming model, so as to obtain a compressed state space for the problem of dynamic programming; and in the compressed state space, using an APO optimization algorithm and a DCS algorithm to construct a time-series production simulation method solution framework taking capacity dynamic programming into consideration, and using the time-series production simulation method solution framework to determine a fast time-series production simulation result taking capacity dynamic programming into consideration.

Description

A time sequence production simulation method considering capacity dynamic programming, device and medium

[0001] The present application claims priority to the Chinese patent application No. 202410891486.6, filed on July 4, 2025, and entitled "A time sequence production simulation method considering capacity dynamic programming, device and medium", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the technical field of power systems, in particular to a time sequence production simulation method considering capacity dynamic programming, device and medium. BACKGROUND

[0003] Under the background of fine management of power system power capacity, there is a locking effect of power capacity, and the power capacity put into production at a certain time will have a great impact on the future. In order to realize the economic benefit and environmental benefit in a long period, it is necessary to plan and design the power system power capacity from a long period perspective. However, there is a problem of difficulty in solving. Dynamic programming problem can efficiently solve problems with overlapping subproblems, can handle multi-stage decision problems, and can obtain optimal solution by recursively and combining the solutions of subproblems, and has fast running time. For the problem of power system power capacity expansion considering capacity retirement, the problem can be modeled as a dynamic programming problem with aftereffect. However, there is a problem of difficulty in solving the dynamic programming problem with aftereffect in the mathematical level. SUMMARY

[0004] The purpose of the present application is to provide a time sequence production simulation method considering capacity dynamic programming, device and medium, which realizes the rapid simulation of time sequence production.

[0005] In order to achieve the above purpose, the present application provides the following solutions:

[0006] In a first aspect, the application provides a time-series production simulation method considering capacity dynamic programming, comprising: establishing a single-year time-series production simulation model; the single-year time-series production simulation model comprises a single-year total cost objective function and a first constraint condition; constructing an equivalent load support rate interval decision network based on the single-year time-series production simulation model and a CNN-GRU network; constructing a multi-stage dynamic programming model for power supply capacity expansion based on the single-year time-series production simulation model; the multi-stage dynamic programming model comprises a dynamic programming objective function and the first constraint condition; using a NO search algorithm to compress a state space in the multi-stage dynamic programming model to obtain a compressed state space of the dynamic programming problem; using an APO optimization algorithm and a DCS algorithm to construct a time-series production simulation method solving framework considering capacity dynamic programming in the compressed state space; and determining a fast time-series production simulation result considering capacity dynamic programming by using the time-series production simulation method solving framework, the equivalent load support rate interval decision network and the multi-stage dynamic programming model.

[0007] In a second aspect, the application provides a computer device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the time-series production simulation method considering capacity dynamic programming.

[0008] In a third aspect, the application provides a computer readable storage medium, which stores a computer program executable by a processor to implement the time-series production simulation method considering capacity dynamic programming.

[0009] In a fourth aspect, the application provides a computer program product, comprising a computer program executable by a processor to implement the time-series production simulation method considering capacity dynamic programming. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0011] FIG. 1 is a flowchart of a time-series production simulation method considering capacity dynamic programming according to an embodiment of the application;

[0012] FIG. 2 is a schematic diagram of a fast time-series production simulation method framework considering capacity dynamic programming;

[0013] FIG. 3 is a process diagram of a fast timing production simulation method considering capacity dynamic programming;

[0014] FIG. 4 is a structural schematic diagram of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0016] The purpose of the present application is to provide a timing production simulation method, device and medium considering capacity dynamic programming, aiming to realize fast simulation of timing production.

[0017] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0018] In an exemplary embodiment, as shown in FIGS. 1-3, the timing production simulation method considering capacity dynamic programming in the present embodiment includes:

[0019] Step 1: Establish a single-year timing production simulation model; the single-year timing production simulation model includes a single-year total cost objective function and a first constraint condition.

[0020] As an optional implementation, the single-year total cost objective function includes: a F inv +F main +F g +F new +F ess (1)

[0021] Wherein, F a is the total cost in a single year; F inv is the investment cost; F main is the maintenance cost; F g is the thermal power operation cost; F new is the wind and light abandoned electricity penalty cost; F ess is the storage energy charging and discharging operation cost; Ω k is a set of each type of unit; a k,y is the annual equivalent parameter of the unit capacity investment cost of the kth type of unit in the yth year; is the rated installed capacity of the kth type of unit in the yth year; b k,yis the annual equivalent parameter of the unit capacity maintenance cost of the kth type of unit in the yth year; Ω s is a set of operation scenarios; Ω t is a set of operation time points in the scenario; is the unit electricity cost coefficient of the thermal power unit in the yth year; p g,s,t is the actual output of the thermal power unit at the tth time point in the s th scenario; Δt is the time interval; is the unit penalty cost coefficient of the wind and photovoltaic unit in the yth year for the abandoned electricity; p is the predicted output of the photovoltaic unit at the tth time point in the s th scenario; p pv,s,t is the actual output of the photovoltaic unit at the tth time point in the s th scenario; is the predicted output of the wind power unit at the tth time point in the s th scenario; p w,s,t is the actual output of the wind power unit at the tth time point in the s th scenario; is the unit power operation cost coefficient of the electrochemical energy storage system in the yth year; p is the actual discharge power of the electrochemical energy storage system; p is the actual charging power of the electrochemical energy storage system.

[0022] As an optional implementation, the first constraint condition includes: a power balance constraint, a spinning reserve constraint, and a unit characteristic constraint.

[0023] The power balance constraint includes:

[0024] Wherein, p load,s,t is the electrical load value at the tth time point in the s th scenario.

[0025] The spinning reserve constraint includes:

[0026] Wherein, R s,t is the positive spinning reserve at the tth time point in the s th scenario; α max is the maximum technical output coefficient of the thermal power unit; B g,s,t is the grid-connected capacity of the thermal power unit at the tth time point in the s th scenario; is the rated capacity of the electrochemical energy storage system; O s,t is the negative spinning reserve at the tth time point in the s th scenario; α min is the minimum technical output coefficient of the thermal power unit; r pv is the positive reserve coefficient of the photovoltaic unit; r w is the positive reserve coefficient of the wind power unit; r load is the positive reserve coefficient of the electrical load; o pv is the negative reserve coefficient of the photovoltaic unit; o w is the negative reserve coefficient of the wind power unit; o loadThe standby coefficient of the electrical load.

[0027] The unit characteristic constraint includes:

[0028] wherein, is the rated capacity of the thermal power unit; p g,s,t-1 is the actual output of the thermal power unit at the t-1 moment in the s scene; β up is the ramp-up rate of the thermal power unit; β down is the ramp-down rate of the thermal power unit; is the rated capacity of the photovoltaic unit; is the rated capacity of the wind power unit; E ess,s,t is the actual energy value of the electrochemical energy storage system at the t moment in the s scene; E ess,s,t-1 is the actual energy value of the electrochemical energy storage system at the t-1 moment in the s scene; η ess is the charge-discharge efficiency coefficient of the electrochemical energy storage system; SoC min is the minimum state of charge of the electrochemical energy storage system; SoC max is the maximum state of charge of the electrochemical energy storage system; is the rated energy of the electrochemical energy storage system; E ess,s,0 is the initial energy of the electrochemical energy storage system in the dispatching cycle; E ess,s,T is the final energy of the electrochemical energy storage system in the dispatching cycle.

[0029] Specifically, the single-year time sequence production simulation model is established based on the scenario method: first, the historical data of wind power output, photovoltaic output and electrical load are obtained, and the k-means clustering algorithm is used to generate typical day scenarios. The typical day scenario specifically represents the wind power output, photovoltaic output and electrical load data in a typical day, and its method is to replace the data of one year with the data of a typical day scenario. Secondly, a single-year time sequence production simulation model is established based on the typical day scenario. Finally, the single-year time sequence production simulation model is established as a pure linear programming problem, which has excellent solvability, and a Python Gurobi solver is used for solving. The input data is the typical day scenario data, the installed capacity of the unit (the unit includes thermal power unit, wind power unit, photovoltaic unit and electrochemical energy storage system), and the output data is the unit operation in a typical day.

[0030] Step 2: Based on the single-year time sequence production simulation model and the CNN-GRU network, an equivalent load support rate interval decision network is constructed.

[0031] As an optional implementation, step 2 includes:

[0032] Step 21: Obtain a preset capacity retirement sequence.

[0033] Step 22: based on the CNN network and the GRU network, a CNN-GRU network is constructed.

[0034] Step 23: based on the equivalent load support rate, a single-year time series production simulation model is reconstructed to obtain a reconstructed single-year time series production simulation model.

[0035] Step 24: within the preset capacity retirement range, a random function is used to constantly generate a capacity retirement sequence in the planning period.

[0036] Step 25: the capacity retirement sequence is input into the reconstructed single-year time series production simulation model to obtain the corresponding equivalent load support rate interval.

[0037] Step 26: using each capacity retirement sequence and the corresponding equivalent load support rate interval as a training data set, the CNN-GRU network is trained to obtain an equivalent load support rate interval decision network.

[0038] Specifically, first, based on the single-year time series production simulation model, the equivalent load support rate interval [δ bot ,δ top ] under the capacity retirement sequence is generated.

[0039] Under a given capacity retirement sequence, the electric load in the single-year time series production simulation model is reconstructed as: p’ load,s,t =δ1p load,s,t (8)

[0040] Where p’ load,s,t is the electric load value at the t time of the s scene after the first reconstruction; δ1 is the equivalent load support rate, which is a decision variable to be optimized. The minimum total cost is taken as the objective to obtain the low value δ bot of the equivalent load support rate at this time; the maximum equivalent load support rate is taken as the objective to obtain the high value δ top of the equivalent load support rate at this time.

[0041] Secondly, a data model hybrid driven method for generating equivalent load support rate interval under capacity retirement sequence is established. Based on the convolutional neural network-gated recurrent unit (CNN-GRU), a data-driven equivalent load support rate interval decision network is established. First, the random function is used to generate the capacity retirement sequence in the planning period, and the capacity retirement sequence is input into the reconstructed model to obtain the equivalent load support rate interval. A large number of capacity retirement sequences and corresponding equivalent load support rate intervals are obtained to construct the training data set. Second, the CNN-GRU network is constructed, and the deep learning model is used to learn and imitate the massive data. The CNN deep learning network shows very excellent performance in feature extraction. The CNN uses local connection and weight sharing, which can directly extract deep feature information of data through convolutional layers and pooling layers alternately. Therefore, CNN is used to fully extract deep time series information to form high-dimensional feature vector data as the mapping input of the subsequent network. Since there is a close time sequence relationship between the capacity retirement sequence, the GRU which is good at processing high-dimensional time sequence feature data and has fast operation speed is used to establish the mapping relationship between the capacity retirement sequence and the equivalent load support rate interval decision. In the CNN-GRU hybrid network, CNN is mainly responsible for feature extraction, and GRU is mainly responsible for establishing the mapping relationship.

[0042] Finally, the Adam algorithm is used to train the CNN-GRU network, and the weight update formula is:

[0043] where ω h+1 is the network weight under the h+1 training times; ω h is the network weight under the h training times; τ is the learning rate; is the second moment mean of the corrected gradient under the h training times; θ is the smoothing coefficient; is the first moment mean of the corrected gradient under the h training times; β2 is the decay factor of the second moment mean; v h-1 is the second moment mean of the gradient under the h-1 training times; is the gradient operator; ω h is the network weight under the h training times; is the h power of β2; β1 is the decay factor of the first moment mean; m h-1 is the first moment mean of the gradient under the h-1 training times; is the h power of β1.

[0044] The CNN-GRU network is trained by using a variable learning rate method, that is, the learning rate decreases with the increase of the number of training times. The loss function of model training is defined as:

[0045] wherein, R MSE is the root mean square error; is the actual value of the h1th equivalent load support rate interval, q h1 is the estimated value of the h1th equivalent load support rate interval output by the CNN-GRU network; and H is the number of equivalent load support rate intervals.

[0046] Step 3: Based on the single-year time series production simulation model, a multi-stage dynamic programming model for the power supply capacity expansion problem is constructed; the multi-stage dynamic programming model includes a dynamic programming objective function and a first constraint condition.

[0047] As an optional implementation, the dynamic programming objective function includes:

[0048] wherein, f k (·) is the dynamic programming objective function of the kth stage; α k is the state of the kth stage; V k is the cumulative target value of the kth stage; x0 is the decision variable of the initial stage; x1 is the decision variable of the first stage; x k is the decision variable of the kth stage.

[0049] Specifically, when solving the power supply capacity expansion problem, the power supply capacity of each stage in the planning period needs to be decided. The multi-stage dynamic programming model takes the power supply capacity expansion problem of the power system as the research content, and uses the dynamic programming method to mathematically represent the power supply capacity expansion problem.

[0050] The objective function of the multi-stage dynamic programming model is the total cost, including the investment cost, maintenance cost, thermal power operation cost, wind and light abandoned power penalty cost and energy storage charging and discharging operation cost in each stage. The state is the installed capacity of each type of power source in each stage. The stage is the time scale of making a power supply capacity decision. The decision is the selection behavior from one state to the next state in the stage, specifically the selected installed capacity of each type of power source for expansion. The strategy is the sequence composed of the decisions of each stage, that is, the strategy of the capacity expansion of each type of power source.

[0051] The power system power supply capacity expansion problem is characterized by the dynamic programming method:

[0052] 1) Stage: The planning period is divided into stages, one year is divided into one stage, and there are Y years, that is, there are Y stages in the planning period, and stage k represents the kth stage.

[0053] 2) State: The feasible power capacity in each stage is represented as state α, which has four types of wind power, photovoltaic, thermal power and electrochemical energy storage, and each state vector dimension is 4. For the state candidate set representation in the stage, a fixed capacity interval is used to generate the state subset I.

[0054] 3) Decision of state transition:

[0055] Decision variable: x k,j (α k,i ) is the decision variable from α k,i to α k+1,j , that is, the capacity of the kth stage expansion, and there is the following relationship: α k+1,j = α k,i +x k,j (α k,i ) (14)

[0056] Where, α k+1,j is the jth state of the k+1th stage; α k,i is the ith state of the kth stage.

[0057] 4) The minimum planning operation cost (investment cost, maintenance cost, thermal power operation cost, wind and light abandoned power penalty cost and energy storage charging and discharging operation cost) is the optimal index function.

[0058] Index function form: V k,j = v k (α k,i ,x k,j )+V k-1,i (16)

[0059] Where, V k,j is the cumulative target value of the jth state of the k+1th stage; v q (α q ,x q ) is the target value of the qth stage; α q is the state of the qth stage; x q is the decision variable of the qth stage; v k (α k,i ,x k,j ) is the target value of the kth stage; V k-1,i is the cumulative target value of the jth state of the k-1th stage.

[0060] 5) The optimal objective function, that is, the dynamic programming objective function is:

[0061] The input data of the multi-stage dynamic programming model is the typical daily scene data and the equivalent load support rate.

[0062] Regarding the solution of the optimal objective function, within a certain state, the planning operation cost is calculated based on the single-year time-series production simulation model in step 1, and is corrected for calculation. The electric load in the corrected single-year time-series production simulation model is reconstructed as: p” load，s，t = (1 + δ2) p load，s，t (17)

[0063] where p” load，s，t is the electric load value at the t time in the s scene after the second reconstruction; δ2 is the equivalent load support rate in the input data of the multi-stage dynamic programming model.

[0064] With the goal of minimizing the planning operation cost, the Python calls the Gurobi solver for solution, and returns the target value of the current state of the multi-stage dynamic programming model.

[0065] Step 4: Use the NO search algorithm to compress the state space of the multi-stage dynamic programming model, and obtain the compressed state space of the dynamic programming problem.

[0066] Specifically, when the stage and state set of the multi-stage dynamic programming model increases, the computer's computing workload and required storage space will increase sharply, forming a "dimension disaster" problem. First, according to the minimum single-year expansion planning model, a locally optimal capacity decision scheme is obtained, and finally a locally optimal search path is retained. The non-monopoly (NO) search algorithm is used to explore and develop other possible more excellent search paths from the locally optimal search path. When searching for a path, take the locally optimal value (the target value under the locally optimal search path) as the benchmark. If the target value of the new search path is worse than the locally optimal value, mark the surrounding area of the new search path region as unusable space, and the degree of marking space is affected by the difference in target values; similarly, if the target value of the new search path is better than the locally optimal value, mark the surrounding area of the new search path region as usable space. Finally, after the search is completed, according to the state space marking result, the advantages and disadvantages are offset, and the state space with better marking result is preferentially selected as the compressed state space of the dynamic programming problem.

[0067] The formula for the search path influence range is:

[0068] where A k,i,n is the position of the i state in the k stage under the n search path, A k,j,0 is the position of the j state in the k stage under the initial locally optimal search path; σ + is the range coefficient of excellent effect; f0 is the locally optimal target value; f n is the target value of the new search path; σ - is the range coefficient of inferior effect.

[0069] If the marking space satisfies the above formula, Ω k-i The marking degree is represented by the following formula:

[0070] Where, z k,i is the marking degree of the marking space of the kth stage and the ith state, and p is the marking degree coefficient.

[0071] According to the marking result after iteration, the state space is selected, and the criterion is defined as: z k,i ≥ ε (20)

[0072] Where, ε is a constant coefficient.

[0073] The non-monopoly search algorithm is a single-solution metaphor-free algorithm that combines the advantages of exploration and development, and can avoid the problem of falling into a suboptimal solution. First, the population is initialized according to the local optimal search path and the capacities of various units in the planning period; Next, according to formula (21), the target value is calculated, and the search direction is constantly improved by judging the size of the target value. When the upper iteration times are more than half, according to formula (22), development is carried out, that is, formula (22) replaces formula (21), and iteration optimization continues; Finally, the NO algorithm constantly adjusts the search direction until the algorithm reaches the upper limit of the iteration times, and the iteration process ends. X new (j) = rand * X (RP) (21) new (j) = X (j) - [X (SRP) * rand] * eps - [X (j) - NO] (22)

[0074] Where, X new is the newly generated capacity decision scheme of each state; rand is a random value between 0 and 1; j is the dimension index of the jth state under the capacity decision scheme; X is the current capacity decision scheme of each state; RP and SRP are both random dimension indexes, between 1 and the maximum dimension; eps is a small value parameter; NO is an adjustment parameter.

[0075] Step 5: In the compressed state space, use the APO optimization algorithm and the DCS algorithm to build a time-series production simulation method solving framework considering capacity dynamic planning.

[0076] Specifically, the time sequence production simulation method solving framework is divided into an outer layer stage and an inner layer stage. The outer layer stage adopts an artificial protozoa optimizer algorithm to generate a capacity retirement sequence and quickly obtain an equivalent load support rate interval according to a mathematical model hybrid driving method. The inner layer adopts a differential creative search algorithm to generate an equivalent load support rate coefficient according to the interval transmitted by the outer layer, and outputs the equivalent load support rate coefficient to a multi-stage dynamic programming model to obtain the cost of the expansion problem. The retirement cost of the retirement problem is obtained according to the equivalent load support rate coefficient and the capacity retirement sequence, and then the cost of the original problem is obtained. The inner layer is output to the outer layer after iteration is completed, and the outer layer is converged until the output of the fast time sequence production simulation result considering the capacity dynamic programming is obtained.

[0077] An artificial protozoa optimizer (APO) optimization algorithm: imitates the survival behavior of protozoa in nature, including foraging, hibernation and reproduction, and has the characteristics of simplicity and effectiveness for optimization problems. Equation (23) simulates the autotrophic behavior, heterotrophic behavior, hibernation behavior and reproduction behavior of protozoa, respectively. In each iteration, the probabilities of the behaviors of protozoa are first calculated according to equation (24) and the appropriate behavior is selected from equation (23) for iteration, and finally it is judged whether the position information of the protozoa is updated according to equation (25).

[0078] In equation (23), is the updated position of the e-th protozoa; W e is the original position of the e-th protozoa; f1 is the foraging factor; W y is the y-th protozoa randomly selected; N is the number of neighboring points; w1 is the weight factor of autotrophic behavior; W γ- is the original position of a protozoa with a ranking index less than e randomly selected from the neighboring points of the k-th protozoa; W γ+ is the original position of a protozoa with a ranking index greater than e randomly selected from the neighboring points of the k-th protozoa; is the Hadamard product; M f is the mapping vector of autotrophic behavior and heterotrophic behavior; W near is the W e nearby position; w2 is the weight factor of heterotrophic behavior; W e-γ is a protozoa selected from the neighboring points of the γ-th protozoa, with a ranking index of e-γ; W e+γ is a protozoa selected from the neighboring points of the γ-th protozoa, with a ranking index of e+γ; W min is the lower bound position of the protozoa position; Rand is a random vector with elements in the interval [0, 1]; W max is the upper bound position of the protozoa position; M ris the mapping vector for reproductive behavior.

[0079] In formula (24), pf is the proportion fraction of dormancy and reproduction in protozoan population; pf max is the maximum value of pf; p ah is the probability of autotrophic and heterotrophic behavior, iter is the current iteration number; iter max is the upper limit of iteration number; p dr is the probability of dormancy and reproduction; ps is the size of protozoan population.

[0080] In formula (25), is the target value of the update position of the e-th protozoan; F(W e ) is the target value of the original position of the e-th protozoan.

[0081] Differential creative search algorithm:

[0082] Differential creative search (DCS) is a breakthrough optimization algorithm that completely changes the traditional decision-making system in complex environments. The main goal of DCS is to improve decision-making efficiency by adopting a dual-strategy method that balances divergent and convergent thinking. DCS method first differentiates knowledge acquisition for randomly initialized decision variables of team members according to formula (26) to determine the degree of imperfect knowledge of team members; formula (27) represents the iteration equations of convergent thinking and divergent thinking, respectively, and further determines the mutation direction of decision variables and obtains updated decision variable information according to formula (27) combined with the current degree of imperfect knowledge of team members; finally, formula (28) and formula (29) are used for decision variable information iteration and obtaining the best target value.

[0083] In formula (26), is the variable value of the l-th member at the t-th iteration; R l,t is the sequence number of the l-th member at the beginning of the t-th iteration; NP is the number of team members.

[0084] In formula (27), o l,d is the updated d-th element of the l-th member; w v is the best cognitive weight; u best,d is the d-th element of the best member in the current iteration; λ l,t and ω l,t are the coefficient values of the l-th member at the t-th iteration; u r2,d is the d-th element of the r2-th member randomly selected from {1, 2,..., NP}; u l,d is the d-th element of the l-th member before updating; u r1,dLk(a) is the dth element of the rth member of {1,2,..., NP} randomly selected; Lk(a v ,σ v ) are control parameters of a Linnik distribution random number generator with ξ v and σ v .

[0085] In equation (28), D l,t+1 is the updated variable value of the lth member at the t+1th iteration; C l,t is the variable value of the lth member before updating at the tth iteration; F2(D l,t ) is the updated target value of the ith member at the tth iteration; F2(C l,t ) is the target value of the ith member before updating at the tth iteration; D l,t is the updated variable value of the lth member at the tth iteration.

[0086] In equation (29), C best,t+1 is the best-performing member at the t+1th iteration; C l,t+1 is the variable value of the lth member before updating at the t+1th iteration; F2(C l,t+1 ) is the target value of the ith member before updating at the t+1th iteration; F2(C best,t ) is the target value of the best-performing member at the tth iteration; C best,t is the best-performing member at the tth iteration.

[0087] Step 6: Determine the fast time series production simulation result considering capacity dynamic programming by using the time series production simulation method to solve the framework, equivalent load support rate interval decision network and multi-stage dynamic programming model.

[0088] The application also provides an application scenario of the time sequence production simulation method considering capacity dynamic programming. Specifically, the time sequence production simulation method considering capacity dynamic programming can be applied in a power system production simulation scenario. In the power system production simulation scenario, a single-year time sequence production simulation model is established; an equivalent load support rate interval decision network is constructed based on the single-year time sequence production simulation model and a CNN-GRU network; a multi-stage dynamic programming model for power source capacity expansion is constructed based on the single-year time sequence production simulation model; the multi-stage dynamic programming model includes a dynamic programming objective function and a first constraint condition; a state space of the multi-stage dynamic programming model is compressed by using a NO search algorithm to obtain a compressed state space of the dynamic programming problem; a time sequence production simulation method considering capacity dynamic programming solving framework is constructed by using an APO optimization algorithm and a DCS algorithm in the compressed state space; and a fast time sequence production simulation result considering capacity dynamic programming is determined by using the time sequence production simulation method solving framework, the equivalent load support rate interval decision network and the multi-stage dynamic programming model.

[0089] In an exemplary embodiment, a computer device is provided, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor executing the computer program to implement the time sequence production simulation method considering capacity dynamic programming in embodiment 1.

[0090] In an exemplary embodiment, a computer readable storage medium is provided, having stored thereon a computer program, the computer program being executable by a processor to implement the time sequence production simulation method considering capacity dynamic programming in embodiment 1.

[0091] In an exemplary embodiment, a computer program product is provided, comprising a computer program, the computer program being executable by a processor to implement the time sequence production simulation method considering capacity dynamic programming in embodiment 1.

[0092] In an exemplary embodiment, a computer device is provided, which can be a server or a terminal, and an internal structure diagram thereof can be as shown in FIG. 4. The computer device includes a processor, a memory, an input / output interface (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 operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is used to store time-series production simulation data considering dynamic capacity planning. 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 network connection. The computer program is executed by the processor to implement a time-series production simulation method considering dynamic capacity planning.

[0093] Those skilled in the art can understand that the structure shown in FIG. 4 is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or less components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0094] It should be noted that the user information (including but not limited to user equipment 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 authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0095] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in each embodiment provided by the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0096] According to the specific embodiments provided by the present application, the following technical effects are disclosed:

[0097] The application discloses a time sequence production simulation method and device considering capacity dynamic programming, a medium and a product, proposes a single-year time sequence production simulation model based on a scenario method, establishes the single-year time sequence production simulation model as a pure linear programming problem, has excellent solvability, and can achieve the purpose of fast solving under a solver or a heuristic algorithm; a data-model hybrid driving method is used to quickly generate an equivalent load support rate interval under a capacity decommissioning sequence, and has the characteristics of being fast and high in precision; the application proposes a multi-stage dynamic programming model establishment method for power supply capacity expansion problems, can use a dynamic programming method to quickly solve the power supply capacity expansion problems, and improves the solving speed; a dynamic programming problem state space compression method based on a NO search algorithm can solve the dynamic programming problem dimension disaster problem, can effectively improve the solving speed for large-scale dynamic programming problems, and ensures high precision; and the fast time sequence production simulation method considering capacity dynamic programming has a solving framework, the solving framework can quickly solve dynamic programming problems with aftereffects, uses an APO optimization algorithm and a DCS algorithm to improve the searching and optimization ability of the algorithm, and has the characteristics of high searching efficiency.

[0098] The database involved in each embodiment provided by the application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, and the like, and is not limited thereto. The processor involved in each embodiment provided by the application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, and the like, and is not limited thereto.

[0099] Each technical feature of the above embodiments can be combined arbitrarily, and to make the description concise, each technical feature in the above embodiments is not described in all possible combinations, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0100] The principles and implementation modes of the application are described by using specific examples in the present application, and the above embodiment description is only used to help understand the method and core idea of the application; meanwhile, for those skilled in the art, according to the idea of the application, the specific implementation mode and application range can be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A time series production simulation method considering capacity dynamic programming, characterized by, The method comprises: establishing a single-year time series production simulation model; the single-year time series production simulation model comprises a single-year total cost objective function and a first constraint condition; based on the single-year time series production simulation model and the CNN-GRU network, an equivalent load support rate interval decision network is constructed; based on the single-year time series production simulation model, a multi-stage dynamic programming model for power supply capacity expansion problems is constructed; the multi-stage dynamic programming model comprises a dynamic programming objective function and the first constraint condition; using the NO search algorithm, the state space in the multi-stage dynamic programming model is compressed to obtain a compressed state space of the dynamic programming problem; in the compressed state space, an APO optimization algorithm and a DCS algorithm are used to construct a time series production simulation method solving framework considering capacity dynamic programming; the time series production simulation method solving framework, the equivalent load support rate interval decision network and the multi-stage dynamic programming model are used to determine the fast time series production simulation result considering capacity dynamic programming.

2. The timing production simulation method considering capacity dynamics planning according to claim 1, wherein, The single-year total cost objective function includes: F a = F inv + F main + F g + F new + F ess ; Wherein, F a is the total cost in a single year; F inv is the investment cost; F main is the maintenance cost; F g is the thermal power operation cost; F new is the wind and light abandoned electricity penalty cost; F ess is the energy storage charging and discharging operation cost; Ω k is each type of unit set; a k,y is the annual equivalent parameter of the unit capacity investment cost of the kth type of unit in the yth year; is the rated installed capacity of the kth type of unit in the yth year; b k,y is the annual equivalent parameter of the unit capacity maintenance cost of the kth type of unit in the yth year; Ω s is the set of each operating scenario; Ω t is the set of each operating time in the scenario; is the cost coefficient of the unit power of the thermal power unit in the yth year; p g,s,t is the actual output of the thermal power unit at the tth moment in the st scenario; Δt is the time interval; For the yth year, the penalty cost coefficient of the unit abandoned electricity of wind and photovoltaic units; Pstis the predicted power output of the photovoltaic unit at time t and scenario s; p pv,s,t Pstis the predicted power output of the photovoltaic unit at time t and scenario s; p Pstis the predicted output of the wind turbine at time t and scenario s; p w,s,t Pstis the predicted output of the wind turbine at time t and scenario s; p coefficient of the unit power operating cost of the electrochemical energy storage system for the yth year; for the actual discharge power of an electrochemical energy storage system; The actual charging power of the electrochemical energy storage system.

3. The timing production simulation method considering capacity dynamics planning according to claim 2, wherein, The first constraint condition comprises a power balance constraint, a spinning reserve constraint and a unit characteristic constraint; The power balance constraint includes: wherein p load,s,t is the electrical load value at time t in the s-th scenario; The rotation standby constraint includes: wherein R s,t is the positive rotation reserve at the tth time in the s th scenario; a max is the maximum technical output coefficient of the thermal power unit; B g,s,t is the grid-connected capacity of the thermal power unit at the tth time in the s th scenario; is the rated capacity of the electrochemical energy storage system; O s,t is the negative spinning reserve of the t-th time in the s-th scenario; a min is the minimum technical output coefficient of the thermal power unit; r pv is the positive reserve coefficient of the photovoltaic unit; r w is the positive reserve coefficient of the wind power unit; r load is the positive reserve coefficient of the electric load; o pv is the negative reserve coefficient of the photovoltaic unit; o w is the negative reserve coefficient of the wind power unit; o load is the negative reserve coefficient of the electric load; The unit characteristics constraints include: wherein P is the rated capacity of the thermal power unit; p g,s,t-1 P is the actual output of the thermal power unit at the t-1 moment in the s scene; β up P is the climbing rate of the thermal power unit; β down P is the descending rate of the thermal power unit; the rated capacity of the photovoltaic unit; E is the rated capacity of the wind turbine; E ess,s,t E is the actual energy value of the electrochemical energy storage system at the t time instant under the s scenario; E ess,s,t-1 E is the actual energy value of the electrochemical energy storage system at the t-1 time instant under the s scenario; E ess η is the charge-discharge efficiency coefficient of the electrochemical energy storage system; η min SoC is the minimum state of charge of the electrochemical energy storage system; SoC max SoC is the maximum state of charge of the electrochemical energy storage system; SoC E is the rated energy of the electrochemical energy storage system ess,s,0 E0is the initial energy of the electrochemical energy storage system at the beginning of the dispatch period ess,s,T E1is the final energy of the electrochemical energy storage system at the end of the dispatch period 4. The timing production simulation method considering capacity dynamics planning according to claim 1, wherein, based on the single-year time series production simulation model and the CNN-GRU network, an equivalent load support rate interval decision network is constructed, comprising: obtaining a preset capacity retirement sequence; based on a CNN network and a GRU network, the CNN-GRU network is constructed; based on the equivalent load support rate, the single-year time series production simulation model is reconstructed to obtain a reconstructed single-year time series production simulation model; within the preset capacity retirement range, a random function is used to constantly generate a capacity retirement sequence in the planning period; the capacity retirement sequence is input into the reconstructed single-year time series production simulation model to obtain a corresponding equivalent load support rate interval; using each of the capacity retirement sequences and the corresponding equivalent load support rate interval as a training data set, the CNN-GRU network is trained to obtain the equivalent load support rate interval decision network.

5. The timing production simulation method considering capacity dynamics planning according to claim 1, wherein, The dynamic programming objective function includes: wherein f k (·) is the dynamic programming objective function of the kth stage; a k is the state of the kth stage; V k is the cumulative objective value of the kth stage; x0is the decision variable of the initial stage; x1is the decision variable of the 1st stage; x k is the decision variable of the kth stage.

6. A computer apparatus comprising: A memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the time series production simulation method considering capacity dynamic programming of any one of claims 1-5.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the time series production simulation method considering capacity dynamic programming of any one of claims 1-5.

8. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the time series production simulation method considering capacity dynamic programming of any one of claims 1-5.

Images