Power-carbon market coupling decision optimization method and device based on probability distribution mapping, terminal equipment and storage medium

Abstract

The invention discloses an electricity-carbon market coupling decision optimization method and device based on probability distribution mapping, terminal equipment and a storage medium, and belongs to the technical field of power dispatching decision optimization, the method comprises the following steps: obtaining current electricity-carbon coupling data, and combining a preset distribution mapping function to construct an electricity-carbon price coupling decision model and constraint conditions; solving an upper-layer model to obtain the output of the generator set, and calculating optimal power-carbon coupling clearing data according to the output of the generator set and a preset distribution mapping function; and according to the optimal power-carbon coupling clearing data and the current power-carbon coupling data, solving the upper-layer model to obtain the optimal power output of the generator set, and scheduling the generator set according to the optimal power output of the generator set. Through the implementation of the method, the problem that the calculation efficiency is low due to the fact that the double-layer game model of the dynamic carbon emission intensity needs to be repeatedly iteratively solved in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to the field of power dispatch decision optimization technology, and in particular to a method, apparatus, terminal equipment and storage medium for coupled decision optimization of the electricity carbon market based on probability distribution mapping. Background Technology

[0002] With the deepening of the coordinated development of my country's electricity and carbon markets, the cost structure of traditional generating units has undergone significant changes, profoundly impacting the bidding behavior of market participants. To maximize the returns of power generators in both the electricity and carbon markets, it is necessary to study the coupled decision-making problems of power generators in the electricity and carbon markets.

[0003] In existing technologies, the corresponding description is usually done by a two-level game model that considers dynamic carbon emission intensity. When performing optimization calculations, the two-level model needs to be solved iteratively, which generates a huge computational burden and leads to low computational efficiency. Summary of the Invention

[0004] This invention provides a method, apparatus, terminal device, and storage medium for coupled decision optimization in the electric carbon market based on probability distribution mapping. It can solve the problem in the prior art that the two-level game model of dynamic carbon emission intensity needs to be iteratively solved repeatedly, resulting in a huge computational burden and thus low computational efficiency.

[0005] One embodiment of the present invention provides a coupled decision optimization method for the electric carbon market based on probability distribution mapping, comprising:

[0006] Obtain the current electrocarbon coupling data;

[0007] Based on current electricity-carbon coupling data, an electricity-carbon price coupling decision model and corresponding constraints are constructed with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading. The electricity-carbon price coupling decision model includes an upper-level model aimed at maximizing generator profits and a lower-level model aimed at minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading. The constraints include a first constraint corresponding to the upper-level model and a second constraint corresponding to the lower-level model.

[0008] Based on the above upper-level model and the current electric carbon coupling data, the generator output that satisfies the above first constraint condition is obtained, and the optimal electric carbon coupling clearing data corresponding to the above lower-level model is calculated based on the above generator output and the preset distribution mapping function.

[0009] Based on the optimal electro-carbon coupling clearing data and the current electro-carbon coupling data, the upper-level model is solved under the first constraint to obtain the optimal generator set output, and the generator sets are scheduled according to the optimal generator set output.

[0010] Furthermore, the construction of the aforementioned preset distribution mapping function includes:

[0011] Acquire historical electro-carbon coupling data within a historical time period;

[0012] Based on the aforementioned historical electro-carbon coupling data, the distribution mapping function construction operation is repeatedly performed until the root mean square error of the current distribution mapping function is not greater than a preset threshold, and the current distribution mapping function is used as the aforementioned preset distribution mapping function.

[0013] The above-mentioned distribution mapping function construction operation includes:

[0014] Obtain historical carbon coupling data for the current historical period, as well as the current optimized output of the generator set; wherein, the initial historical carbon coupling data is the historical carbon coupling data corresponding to the initial historical period, and the initial optimized output of the generator set is the preset initial optimized output of the generator set;

[0015] Based on historical electrocarbon coupling data for the current historical period, the current distribution mapping function is fitted to obtain the current distribution mapping function.

[0016] Based on the current optimized output of the generator set, the current distribution mapping function, and the actual carbon coupling clearing data in the historical carbon coupling data of the current historical period, the root mean square error of the current distribution mapping function is calculated.

[0017] When the root mean square error of the current distribution mapping function exceeds the aforementioned preset threshold, based on the historical electricity-carbon coupling data for the current historical period, and with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading, a coupled decision model for electricity-carbon prices and corresponding current historical constraints are constructed for the current historical period. The coupled decision model for electricity-carbon prices for the current historical period includes: a current historical upper-level model with the objective of maximizing generator profits, and a current historical lower-level model with the objectives of minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading. The current historical constraints include: a first current historical constraint corresponding to the aforementioned current historical upper-level model, and a second current historical constraint corresponding to the current historical lower-level model.

[0018] Based on the current optimized output of the generator set, the current historical lower-level model is solved under the second current historical constraint to obtain the current electric carbon coupling clearing optimization data;

[0019] Based on the current electricity-carbon coupling clearing optimization data and the historical electricity-carbon coupling data for the current historical period, the electricity-carbon price coupling decision model for the current historical period is solved under the current historical constraints to obtain the optimized output of the generator units for the next historical period.

[0020] Furthermore, the current distribution mapping function, obtained by fitting historical electrocarbon coupling data for the current historical period, includes:

[0021] The historical carbon dioxide depletion response parameters corresponding to the current historical period are extracted from the historical carbon dioxide coupling data of the current historical period.

[0022] Based on the historical carbon-electric coupling data for the current historical period, the probability density function for the current historical period is calculated.

[0023] Based on the historical carbon clearing response parameters corresponding to the current historical period, the joint probability density function for the current historical period is calculated.

[0024] The distribution difference quantification index value for the current historical period is calculated based on the probability density function of the current historical period and the joint probability density function of the current historical period.

[0025] Based on the preset time basis function, the preset space basis function, and the historical carbon clearing response parameters corresponding to the current historical period, the coefficients of the initial dynamic weight matrix corresponding to the current historical period are fitted and calculated.

[0026] The dynamic weight matrix coefficients corresponding to the current historical period are calculated based on the initial dynamic weight matrix coefficients corresponding to the current historical period and the distribution difference quantification index value of the current historical period.

[0027] Based on the dynamic weight matrix coefficients corresponding to the current historical period, the preset time basis function, and the preset spatial basis function, the current distribution mapping function is constructed.

[0028] Furthermore, based on the dynamic weight matrix coefficients corresponding to the current historical period, the preset time basis function, and the preset spatial basis function, the current distribution mapping function is constructed, including:

[0029] Obtain the rated output of each generator set;

[0030] The memory decay matrix is ​​calculated based on the rated output of each generator set and the corresponding preset carbon emission parameters.

[0031] The actual carbon emission intensity of each unit in the current historical period is extracted from the historical carbon-electric coupling data of the current historical period.

[0032] Based on the memory decay matrix, the dynamic weight matrix coefficients corresponding to the current historical period, the preset time basis function, and the preset space basis function, a distribution mapping function to be fitted for the current historical period is constructed; wherein, the above distribution mapping function to be fitted has a fractional residual term;

[0033] Based on the actual carbon emission intensity of each unit in the current historical period, the distribution mapping function to be fitted for the current historical period is fitted, and several weight coefficients of the distribution mapping function to be fitted for the current historical period are determined.

[0034] Based on the fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random perturbation term, preset time basis function, and preset spatial basis function, the current distribution mapping function is constructed.

[0035] Furthermore, based on the aforementioned fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random perturbation term, preset time basis function, and preset spatial basis function, the current distribution mapping function is constructed, including:

[0036] By taking the derivative of the fractional residual term, we obtain the derivative of the fractional residual term.

[0037] By using numerical integration and the aforementioned derivatives to calculate the integral term of the fractional residual term, an approximate value of the fractional residual term is obtained.

[0038] Based on the above approximation, derivative, fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random disturbance term, preset time basis function, and preset space basis function, construct the linear equation of the above fractional residual term;

[0039] Solving the above linear equation yields the solution for the fractional residual term.

[0040] Based on the solution of the fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random perturbation term, preset time basis function, and preset spatial basis function, the current distribution mapping function is constructed.

[0041] Furthermore, based on the current optimized output of the generator set, the current distribution mapping function, and the actual electric carbon coupling clearing data in the historical electric carbon coupling data for the current historical period, the root mean square error of the current distribution mapping function is calculated, including:

[0042] Based on the current optimized output of the generator set and the current distribution mapping function, the current predicted data for the decarbon coupling of electricity is calculated.

[0043] Based on the current electro-carbon coupling clearing prediction data and the actual electro-carbon coupling clearing data, the root mean square error of the current distribution mapping function is calculated.

[0044] Furthermore, the aforementioned lower-level model includes: a first lower-level model and a second lower-level model;

[0045] The objective function of the above upper-level model is:

[0046]

[0047] In the formula, This represents the profit of power generator g in the electricity market. This indicates the revenue that generator g receives in the carbon market. This indicates the cost of generator g in the carbon market;

[0048] The objective function of the first lower-level model mentioned above is:

[0049]

[0050] In the formula, This represents the bid price of generator set i in bidding block v. This represents the output power of generator unit i in bidding block v during time period t;

[0051] The objective function of the second lower-level model mentioned above is:

[0052]

[0053] In the formula, This represents the carbon market purchase price for generator set i in the bidding block. This represents the amount of carbon allowance purchased by generator set i. This represents the selling price of generator set i in the carbon market within the bidding block. This indicates the amount of carbon quota sold by generator set i.

[0054] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments;

[0055] This invention provides an electric carbon market coupled decision optimization device based on probability distribution mapping, comprising:

[0056] The system includes an electric carbon coupling data acquisition module, a model building module, an optimal electric carbon coupling clearing data calculation module, and a generator set scheduling module.

[0057] The aforementioned electro-carbon coupling data acquisition module is used to acquire the current electro-carbon coupling data;

[0058] The aforementioned model building module is used to construct an electricity-carbon price coupling decision model and corresponding constraints based on current electricity-carbon coupling data, with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading. The electricity-carbon price coupling decision model includes an upper-level model with the objective of maximizing generator profits and a lower-level model with the objectives of minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading. The constraints include a first constraint corresponding to the upper-level model and a second constraint corresponding to the lower-level model.

[0059] The above-mentioned optimal electric carbon coupling clearing data calculation module is used to obtain the generator output that satisfies the above-mentioned first constraint condition based on the above-mentioned upper-level model and the current electric carbon coupling data, and to calculate the optimal electric carbon coupling clearing data corresponding to the above-mentioned lower-level model based on the above-mentioned generator output and the preset distribution mapping function.

[0060] The generator set scheduling module is used to solve the upper-level model under the first constraint condition based on the optimal electric carbon coupling clearing data and the current electric carbon coupling data to obtain the optimal generator set output, and to schedule the generator sets according to the optimal generator set output.

[0061] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment;

[0062] The present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-described method for coupled decision optimization of the electric carbon market based on probability distribution mapping, according to any embodiment of the present invention.

[0063] Based on the above method embodiments, the present invention provides a corresponding storage medium embodiment;

[0064] The present invention provides a storage medium including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-described method for coupled decision optimization of the electric carbon market based on probability distribution mapping, according to any embodiment of the present invention.

[0065] The embodiments of the present invention have the following beneficial effects:

[0066] This invention provides a method, apparatus, terminal device, and storage medium for optimizing coupled decision-making in the carbon market based on probability distribution mapping. The method includes: acquiring current carbon market coupling data; constructing a carbon market price coupled decision-making model and corresponding constraints based on the current carbon market coupling data and a preset distribution mapping function, with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading; obtaining the generator output that satisfies the first constraint based on the upper-level model and the current carbon market coupling data; calculating the optimal carbon market coupling clearing data corresponding to the lower-level model based on the generator output and the preset distribution mapping function; and finally solving the upper-level model under the first constraint based on the optimal carbon market coupling clearing data and the current carbon market coupling data to obtain the optimal generator output, and dispatching the generators based on the optimal generator output. Therefore, this invention uses a preset distribution mapping function to associate the upper and lower models in the coupled decision-making model of electricity and carbon prices. This allows the optimal electricity and carbon price clearing data that originally needed to be substituted into the lower model to be calculated directly during the solution process. There is no need to call the lower model for the solution process. This transforms the two-layer game that requires repeated iterations into a single-layer optimization problem, omitting the interaction process between the upper and lower layers, thus improving computational efficiency. Attached Figure Description

[0067] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0068] Figure 1 This is a flowchart illustrating a probability distribution mapping-based coupled decision optimization method for the electric carbon market, provided by an embodiment of the present invention.

[0069] Figure 2 This is a schematic diagram of a carbon market coupling decision optimization device based on probability distribution mapping, provided in an embodiment of the present invention. Detailed Implementation

[0070] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0071] See Figure 1To address the problem that existing technologies require iterative solutions to a two-layer game model of dynamic carbon emission intensity, resulting in a huge computational burden and low computational efficiency, an embodiment of this invention provides a coupled decision optimization method for the electricity-carbon market based on probability distribution mapping, comprising:

[0072] Step S101: Obtain the current electro-carbon coupling data;

[0073] Specifically, the electricity-carbon coupling data includes: the marginal electricity price of a node, the segmented generation cost of a node, the carbon market clearing price, the carbon quota power capacity of generating units, the carbon quota demand capacity of generating units, the carbon quota selling price under the carbon market stabilization reserve mechanism, the carbon quota buying price under the carbon market stabilization reserve mechanism, the benchmark value of carbon emission intensity of generating units, the actual value of carbon emission intensity of generating units, the maximum trading capacity of carbon quotas of generating units, the binary variable of carbon trading direction of generating units, the load demand of a node, the total output power limit of generating units, the lower limit and upper limit of the power decline rate of generating units, and the maximum transmission capacity of the line.

[0074] Step S102: Based on the current electricity-carbon coupling data, with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading, an electricity-carbon price coupling decision model and corresponding constraints are constructed. The electricity-carbon price coupling decision model includes an upper-level model with the objective of maximizing generator profits, and a lower-level model with the objectives of minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading. The constraints include a first constraint corresponding to the upper-level model and a second constraint corresponding to the lower-level model.

[0075] Specifically, the above-mentioned electricity-carbon price coupling decision model, without the introduction of a preset distribution function, can be simply described as follows:

[0076]

[0077] G(α,y)≤0

[0078] In the formula, α represents the decision variable of the upper-level model, y represents the decision variable of the lower-level model, F(α,y) represents the objective function of the upper-level model, f(α,y) represents the objective function of the lower-level model, G(α,y) represents the first constraint of the upper-level model, and g(α,y) represents the second constraint of the lower-level model.

[0079] Specifically, for the upper-level model, the decision variable is: the output power of generator unit i in bidding block v during time period t.

[0080] Specifically, for the first lower-level model in the lower-level model, its decision variables are: the Lagrange multipliers related to the power balance constraints mentioned above. This parameter also represents the marginal electricity price of node n during time period t, and the total output power of the generator set. For the second lower-level model in the lower-level model, the decision variable is: the purchase price of the carbon market corresponding to generator unit i in the bidding block. The selling price of generator set i in the carbon market within the bidding block Carbon allowances sold by generator set i And the amount of carbon allowance purchased by generator set i

[0081] Specifically, the electricity-carbon price coupling decision-making model before the introduction of a preset distribution function is a multi-time-period, two-level optimization problem. Due to the complex interactions between the upper and lower levels, traditional algorithms struggle to solve it. Therefore, this invention, based on historical decision-making information of market participants, designs a distribution difference quantification index and a dynamic weight adjustment mechanism to construct a distribution mapping function to quantify the impact of upper-level decisions on lower-level decisions. This provides a more efficient and accurate approach to solving the aforementioned model. Based on the distribution mapping function, the model can be reconstructed as follows:

[0082]

[0083] g(α,y)≤0,G(α,y)≤0

[0084] In the formula, f(α) t ,y t α is the correlation function between the upper-level decision and the lower-level response, representing the physical relationship between the decision variables of the upper-level model and the decision variables of the lower-level model at time t. t Let y represent the decision variable of the upper-level model during time period t. t This represents the decision variables of the lower-level model during time period t. This represents the distribution mapping function corresponding to time period t.

[0085] Specifically, the distribution mapping function integrates spatial and temporal characteristics through a linear combination of basis functions and introduces a bias correction mechanism to improve accuracy, which can quantitatively represent the relationship between upper-level decisions and lower-level responses.

[0086] In a preferred embodiment, the construction of the aforementioned preset distribution mapping function includes:

[0087] Acquire historical electro-carbon coupling data within a historical time period;

[0088] Based on the aforementioned historical electro-carbon coupling data, the distribution mapping function construction operation is repeatedly performed until the root mean square error of the current distribution mapping function is not greater than a preset threshold, and the current distribution mapping function is used as the aforementioned preset distribution mapping function.

[0089] Specifically, the aforementioned preset threshold is set to 10. -4 .

[0090] The above-mentioned distribution mapping function construction operation includes:

[0091] Obtain historical carbon coupling data for the current historical period, as well as the current optimized output of the generator set; wherein, the initial historical carbon coupling data is the historical carbon coupling data corresponding to the initial historical period, and the initial optimized output of the generator set is the preset initial optimized output of the generator set;

[0092] Specifically, historical hourly data on carbon-electric coupling over the past year is acquired, and data preprocessing is performed using a rolling window to normalize the data to the [0,1] interval. The time interval is set to 1 hour, meaning the time interval between the current historical period and the next historical period is 1 hour. Therefore, the aforementioned initial historical period is the first historical period in the one-year hourly timeframe. Preferably, data normalization can improve the convergence speed of the overall iterative process, significantly shortening the iterative solution time.

[0093] Based on historical electrocarbon coupling data for the current historical period, the current distribution mapping function is fitted to obtain the current distribution mapping function.

[0094] Based on the current optimized output of the generator set, the current distribution mapping function, and the actual carbon coupling clearing data in the historical carbon coupling data of the current historical period, the root mean square error of the current distribution mapping function is calculated.

[0095] Specifically, the electricity-carbon coupling clearing data refers to the marginal electricity price at node n during time period t. Output power of generator unit i in bidding block v during time period t The carbon market purchase price of generator set i in the bidding block The selling price of generator set i in the carbon market within the bidding block Carbon allowances sold by generator set i And the amount of carbon allowance purchased by generator set i

[0096] When the root mean square error of the current distribution mapping function exceeds the aforementioned preset threshold, based on the historical electricity-carbon coupling data for the current historical period, and with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading, a coupled decision model for electricity-carbon prices and corresponding current historical constraints are constructed for the current historical period. The coupled decision model for electricity-carbon prices for the current historical period includes: a current historical upper-level model with the objective of maximizing generator profits, and a current historical lower-level model with the objectives of minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading. The current historical constraints include: a first current historical constraint corresponding to the aforementioned current historical upper-level model, and a second current historical constraint corresponding to the current historical lower-level model.

[0097] Based on the current optimized output of the generator set, the current historical lower-level model is solved under the second current historical constraint to obtain the current electric carbon coupling clearing optimization data;

[0098] Specifically, based on the current optimized output of the generator set, the current historical lower-level model is solved to obtain the electric carbon coupling clearing optimization data that satisfies the second current historical constraint condition, which is used as the aforementioned current electric carbon coupling clearing optimization data.

[0099] Based on the current electricity-carbon coupling clearing optimization data and the historical electricity-carbon coupling data for the current historical period, the electricity-carbon price coupling decision model for the current historical period is solved under the current historical constraints to obtain the optimized output of the generator units for the next historical period.

[0100] Preferably, when the root mean square error of the current distribution mapping function is greater than a preset threshold, the order of the basis functions can be adjusted, the type of basis functions can be changed, and the optimization of the integration domain can be refined to obtain a more suitable distribution mapping function.

[0101] In this preferred embodiment, a preset distribution mapping function is obtained by repeatedly performing the distribution mapping function construction operation using historical carbon coupling data within a historical time period.

[0102] In another preferred embodiment, the above-mentioned fitting of the current distribution mapping function based on historical electrocarbon coupling data for the current historical period includes:

[0103] The historical carbon dioxide depletion response parameters corresponding to the current historical period are extracted from the historical carbon dioxide coupling data of the current historical period.

[0104] Specifically, the historical carbon clearing response parameters are the results of the lower-level market clearing actually triggered by the upper-level decision during the corresponding historical period. Specifically, they are: the marginal electricity price of node n in time period t, the total output power of the generating units, and the "net carbon allowance surplus or net deficit associated with the generators" in the carbon market.

[0105] Based on the historical carbon-electric coupling data for the current historical period, the probability density function for the current historical period is calculated.

[0106] Specifically, first, based on the decision variables of the upper-level model in the historical carbon-electric coupling data of the current historical period, the corresponding mean vector and covariance matrix are calculated. Then, the probability density function is calculated based on the mean vector and covariance matrix. The formula for calculating the mean vector is:

[0107]

[0108] In the formula, μ α Let represent the mean vector corresponding to the decision variables of the historical layer-level model, nn represent the total number of decision variables of the historical layer-level model, and α represent the mean vector. j Let j represent the j-th decision variable in the historical hierarchical model;

[0109] Specifically, the covariance matrix is ​​calculated using the following formula:

[0110]

[0111] In the formula, Σ α This represents the covariance matrix corresponding to the decision variables of the historical layer model.

[0112] Specifically, the probability density function is calculated using the following formula:

[0113]

[0114] In the formula, f α (α) represents the probability density function corresponding to the decision variables of the historical high-level model, h α This represents the kernel bandwidth corresponding to the probability density function.

[0115] Based on the historical carbon clearing response parameters corresponding to the current historical period, the joint probability density function for the current historical period is calculated.

[0116] Specifically, the joint probability density function is calculated using the following formula:

[0117]

[0118] In the formula, f λ,P,A (λ,P,A) represents the joint probability density function, h λ hP and h A λ represents the different kernel bandwidths in the joint probability density function. j P represents the marginal electricity price in the electricity market corresponding to the j-th group of decision variables in the historical hierarchical model. Specifically, it represents the marginal electricity price at node n in time period t after electricity market clearing. j A represents the actual output of the generating units corresponding to the j-th decision variable in the historical hierarchical model, specifically the total output power of the generating units after the electricity market clearing. j Let A represent the net correlation of the generator's carbon allowances with the j-th group of decision variables in the historical hierarchical model. Specifically, it represents the net surplus or deficit of the generator's carbon allowances in the carbon market. A represents the net correlation of the generator's carbon allowances, P represents the actual total output of the generating units, and λ represents the marginal electricity price in the electricity market.

[0119] The distribution difference quantification index value for the current historical period is calculated based on the probability density function of the current historical period and the joint probability density function of the current historical period.

[0120] Specifically, based on Mahalanobis distance and KL divergence, a quantitative index of distribution difference is obtained, calculated using the following formula:

[0121]

[0122] In the formula, Θ(α,t) represents the quantitative index value of the distribution difference corresponding to the decision variables of the intermediate model in the historical time period t. KL(f) represents the Mahalanobis distance, used to quantify the deviation of a decision variable from its corresponding mean vector. α ||f λ,A,P The ) represents the KL divergence, used to quantify the degree of difference in distribution between the decision variables of the upper-level model and the responses of the lower-level model in history. The smaller the value of the distribution difference quantification index, the more stable the association between the upper-level decisions and the lower-level responses.

[0123] Based on the preset time basis function, the preset space basis function, and the historical carbon clearing response parameters corresponding to the current historical period, the coefficients of the initial dynamic weight matrix corresponding to the current historical period are fitted and calculated.

[0124] Specifically, based on the calculation results of the aforementioned quantitative indicators of distribution differences, it is necessary to further screen and strengthen effective correlation patterns. When the market is stable, the majority of historical correlation patterns should be retained; when the market fluctuates, stable scenario patterns should be prioritized to improve robustness. Therefore, this invention further proposes a dynamic weight adjustment mechanism. By designing weight matrix coefficients containing time-varying functions, the contribution of the basis functions is adaptively adjusted, enabling the correlation patterns between upper-level decisions and lower-level responses to dynamically adapt to the actual market conditions.

[0125] Specifically, with the goal of minimizing the historical response fitting error, the coefficients of the aforementioned initial dynamic weight matrix are calibrated using the least squares method:

[0126]

[0127] In the formula, V represents the coefficients of the initial dynamic weight matrix. true (α t (,t) represents the historical carbon clearing response parameters for time period t. The normalization constraint is represented to ensure the interpretability of the weights. K represents the total number of spatial basis functions, q represents the index of the spatial basis function, l represents the index of the time basis function, and L represents the total number of time basis functions. Describes the spatial basis functions. This represents the time basis function.

[0128] The dynamic weight matrix coefficients corresponding to the current historical period are calculated based on the initial dynamic weight matrix coefficients corresponding to the current historical period and the distribution difference quantification index value of the current historical period.

[0129] Specifically, after determining the initial dynamic weight matrix coefficients, a triple time-varying factor is introduced to dynamically adjust the weights. This triple time-varying factor consists of a distribution difference adjustment factor, a time decay factor, and a time period factor. Finally, combining the initial dynamic weight coefficients with the triple time-varying factors, the final expression for the dynamic weight matrix coefficients is obtained as follows:

[0130]

[0131] In the formula, Θ represents the distribution difference adjustment factor, ∈ represents the sensitivity coefficient, and Θ max The maximum value of the quantitative indicator representing the historical distribution difference. Let t0 represent the time decay factor, t0 represent the start time of all historical periods, and τ represent the decay constant. δ represents the periodicity factor, and δ represents the adjustment coefficient.

[0132] Based on the dynamic weight matrix coefficients corresponding to the current historical period, the preset time basis function, and the preset spatial basis function, the current distribution mapping function is constructed.

[0133] In this preferred embodiment, the current distribution mapping function is obtained by fitting historical carbon coupling data for the current historical period.

[0134] In another preferred embodiment, the above-mentioned construction of the current distribution mapping function based on the dynamic weight matrix coefficients corresponding to the current historical time period, the preset time basis function, and the preset spatial basis function includes:

[0135] Obtain the rated output of each generator set;

[0136] The memory decay matrix is ​​calculated based on the rated output of each generator set and the corresponding preset carbon emission parameters.

[0137] Specifically, to ensure that the dynamic equations of the above fractional residual terms are adapted to the historical characteristics of different generator sets, the memory decay matrix needs to be determined. The memory decay matrix is ​​calculated using the following formula:

[0138]

[0139] In the formula, M represents the memory decay matrix, (m1,m2,…,m N ) represents the element corresponding to each generator unit in the memory decay matrix, m i b represents the element corresponding to generator set i in the memory decay matrix. i The linear coefficient representing the dynamic carbon emission intensity of generator unit i is a parameter related to the carbon emissions of generator unit i. This indicates the rated output of generator set i.

[0140] The actual carbon emission intensity of each unit in the current historical period is extracted from the historical carbon-electric coupling data of the current historical period.

[0141] Based on the memory decay matrix, the dynamic weight matrix coefficients corresponding to the current historical period, the preset time basis function, and the preset space basis function, a distribution mapping function to be fitted for the current historical period is constructed; wherein, the above distribution mapping function to be fitted has a fractional residual term;

[0142] Specifically, the distribution mapping function to be fitted is the distribution mapping function after introducing the fractional residual term. The framework expression of the distribution mapping function before introducing the fractional residual term is:

[0143]

[0144] In the formula, w ql (Θ,t) represents the weight matrix coefficients of the value function in time period t, and l(t) represents the deviation correction term in time period t.

[0145] The specific expression for the introduced distribution mapping function is as follows:

[0146]

[0147] In the formula, Δ γ (t) represents the fractional residual term corresponding to time period t, c l This represents the time basis function φ after expanding the residual term into polynomial form. lThe coefficient of (τ), where l represents the index number of the time basis function.

[0148] Specifically, the dynamic equation expression for the above fractional residual term is:

[0149]

[0150] In the formula, C defines the residual term. Using Caputo fractional derivatives, 0 indicates that the starting point for the integration of the explicit fractional derivative is the initial time, and k p v represents the weighting coefficient corresponding to the historical residuals. qq ξ represents the weighting coefficient corresponding to the second derivative. r Let represent the weighting coefficients corresponding to the third derivative, ∈(t) represent the random disturbance term used to adapt to small market fluctuations, p represent the time interval index of the historical residual term, qq represent the time interval index of the second derivative of the residual term, r represent the time interval index of the third derivative of the residual term, and Δ″ γ (t) represents the second derivative of the fractional residual term corresponding to time period t, Δ″′ γ (t) represents the third derivative of the fractional residual term corresponding to time period t.

[0151] Based on the actual carbon emission intensity of each unit in the current historical period, the distribution mapping function to be fitted for the current historical period is fitted, and several weight coefficients of the distribution mapping function to be fitted for the current historical period are determined.

[0152] Specifically, based on actual carbon emission intensity, the least squares method is used to adjust the weighting coefficient "k". p v qq and ξ r "The goal of fitting the data is to minimize the fitting error of the actual carbon emission intensity, while ensuring that the weighting coefficients after fitting meet the following criteria." To ensure the interpretability of the weighting coefficients, we obtain the specific values ​​of the three weighting coefficients.

[0153] The current distribution mapping function is constructed based on several weight coefficients, the memory decay matrix, the dynamic weight matrix coefficients corresponding to the current historical period, the preset random perturbation term, the preset time basis function, and the preset spatial basis function.

[0154] Specifically, the aforementioned preset random disturbance term, used to adapt to the small fluctuations in the carbon market and the electricity market, is determined by the following formula:

[0155] ∈(t)~N(0,10 -4 )

[0156] Specifically, the final expression of the current distribution mapping function is the same as the expression of the distribution mapping function after introducing the fractional residual term.

[0157] In this preferred embodiment, the current distribution mapping function is constructed based on the dynamic weight matrix coefficients corresponding to the current historical period, the preset time basis function, and the preset spatial basis function.

[0158] In another preferred embodiment, the current distribution mapping function is constructed based on the fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random perturbation term, preset time basis function, and preset spatial basis function, including:

[0159] By taking the derivative of the fractional residual term, we obtain the derivative of the fractional residual term.

[0160] Specifically, we first define the Caputo fractional derivative through nonlocal integration to quantify the impact of historical moments on the current residual term. Its expression is as follows:

[0161]

[0162] In the formula, Γ represents the gamma function, τ' represents the historical moment, τ′∈[0,t], (t-τ') γ This represents the weight decay term; the larger the time interval t-τ', the smaller the weight.

[0163] Subsequently, the second derivative of the fractional residual term is taken to capture the historical characteristics of the residual rate of change, resulting in the aforementioned derivative, where the second derivative of the fractional residual term is obtained using the following formula:

[0164]

[0165] In the formula, φ l " T (τ′) represents the second derivative of the time basis function, used to characterize the time-varying nature of the residual rate of change, Δ″ γ (τ′) represents the second derivative of the fractional residual term corresponding to the historical time τ.

[0166] By using numerical integration and the aforementioned derivatives to calculate the integral term of the fractional residual term, an approximate value of the fractional residual term is obtained.

[0167] Specifically, the Gauss-Legend de Gauss numerical integration method is used to calculate the integral term and obtain an approximate value, which is obtained by the following formula:

[0168]

[0169] In the formula, n I This represents the total number of integration points. In this invention, the total number of integration points is set to 20, τ. qw represents the q-th integration point within the integration interval [0, t]. q This represents the integral weight.

[0170] Based on the above approximation, derivative, fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random disturbance term, preset time basis function, and preset space basis function, construct the linear equation of the above fractional residual term;

[0171] Specifically, substituting the calculated results of the derivative and the memory decay matrix into the dynamic equation of the fractional residual term, and through linearization, we obtain the following system of linear equations for the fractional residual term:

[0172] A·Δ γ (t)=B

[0173] In the formula, A represents the coefficient matrix, which consists of M, k p v qq ξ r The integral weights of the Caputo derivative are used to construct the constant term, which is composed of the historical residuals Δ. γ (tp), second derivative term Δ′ γ ′(t-qq), third derivative term Δ′ γ It consists of ″(tr) and random perturbation terms ∈(t).

[0174] Solving the above linear equation yields the solution for the fractional residual term.

[0175] Specifically, Gaussian elimination is used to solve the linear equations to obtain the solution of the fractional residual terms.

[0176] Based on the solution of the fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random perturbation term, preset time basis function, and preset spatial basis function, the current distribution mapping function is constructed.

[0177] Preferably, after solving the above linear equation system using Gaussian elimination to obtain the numerical solution of the fractional residual term, the subsequent steps are similar to the traditional Galerkin approximation method: First, based on the time dynamics and spatial distribution characteristics of the distribution mapping function, suitable time basis functions and spatial basis functions are selected (i.e., the aforementioned preset time basis functions and preset spatial basis functions, where the preset spatial basis function has an order of 3 and the preset time basis function has an order of 2), and the orthogonality of the basis functions is verified; the distribution mapping function is expanded into a linear combination of basis functions to construct the residual equation; the residual is multiplied by the basis functions to construct a linear equation system, and its coefficients are solved using the orthogonality of the basis functions; finally, the residual error is evaluated, and the approximation function is made to meet the accuracy requirements by adjusting the order of the basis functions, changing the selected basis function types, or subdividing the original integration domain, thus constructing the current distribution mapping function.

[0178] Preferably, while the original distribution mapping function can describe the complex coupling relationship between upper-level decisions and lower-level responses, its framework is still based on integer-order basis functions and cannot describe the historical data dependence characteristics of dynamic parameters in the carbon market. To address this issue, the present invention proposes a fractional-order Galerkin-based distribution mapping function approximation technique. This technique quantifies the impact of historical data on dynamic carbon emission intensity through fractional-order residual terms, effectively improving the approximation accuracy of the function.

[0179] Preferably, the historical dependence characteristics of carbon market parameters are described using Caputo fractional derivatives, and the traditional thousand-level sampling simulation is replaced by Gauss-Legendal numerical integration method (only 20 integration points), which greatly reduces the number of variables and improves computational efficiency.

[0180] In this preferred embodiment, the current distribution mapping function is constructed based on the fractional residual term, several weight coefficients, memory decay matrix, dynamic weight matrix coefficients corresponding to the current historical period, preset random perturbation term, preset time basis function, and preset spatial basis function.

[0181] In another preferred embodiment, the root mean square error of the current distribution mapping function, calculated based on the current optimized generator output, the current distribution mapping function, and the actual carbon coupling clearing data from the historical carbon coupling data of the current historical period, includes:

[0182] Based on the current optimized output of the generator set and the current distribution mapping function, the current predicted data for the decarbon coupling of electricity is calculated.

[0183] Specifically, by substituting the current optimized output of the generator set into the current distribution mapping function, the current predicted data for the carbon-electric coupling cleanup can be calculated.

[0184] Based on the current electro-carbon coupling clearing prediction data and the actual electro-carbon coupling clearing data, the root mean square error of the current distribution mapping function is calculated.

[0185] Specifically, the root mean square error is the root mean square error between the current electro-carbon coupling clearing prediction data calculated by the distribution mapping function and the actual electro-carbon coupling clearing data in the corresponding historical period. The calculation method of this root mean square error is existing technology and will not be elaborated here.

[0186] In this preferred embodiment, the root mean square error of the current distribution mapping function is calculated using the current optimized generator output, the current distribution mapping function, and the actual carbon coupling clearing data in the historical carbon coupling data of the current historical period.

[0187] In another preferred embodiment, the lower-level model includes: a first lower-level model and a second lower-level model;

[0188] The objective function of the above upper-level model is:

[0189]

[0190] In the formula, This represents the profit of power generator g in the electricity market. This indicates the revenue that generator g receives in the carbon market. This indicates the cost of generator g in the carbon market;

[0191] The objective function of the first lower-level model mentioned above is:

[0192]

[0193] In the formula, This represents the bid price of generator set i in bidding block v. This represents the output power of generator unit i in bidding block v during time period t;

[0194] The objective function of the second lower-level model mentioned above is:

[0195]

[0196] In the formula, This represents the carbon market purchase price for generator set i in the bidding block. This represents the amount of carbon allowance purchased by generator set i. This represents the selling price of generator set i in the carbon market within the bidding block. This indicates the amount of carbon quota sold by generator set i.

[0197] In this preferred embodiment, an upper-level model aimed at maximizing the profits of power generators and a lower-level model aimed at minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading are constructed based on current electricity-carbon coupling data.

[0198] Specifically, the aforementioned first constraint includes: market profit equation constraint, generator behavior constraint, and generator carbon quota accounting and scope constraint. The aforementioned market profit equation constraint is as follows:

[0199]

[0200] In the formula, v represents the bidding block. This represents the marginal electricity price of node n during time period t. Let g represent the set of generators on node n. This represents the segmented power generation cost of generator set i in bidding block c. λ represents the output power of generator unit i in bidding block v during time period t, Δt represents the time interval, and λ represents the output power of generator unit i in bidding block v during time period t. C The price represents the carbon market clearing price, where h represents... λ represents the amount of carbon allowance sold by generator set i. FS This indicates the selling price of carbon allowances under the carbon market stabilization reserve mechanism. Let k represent the carbon quota supply capacity of generator set i. λ represents the amount of carbon allowance purchased by generator set i. FB This indicates the purchase price of carbon allowances under the carbon market stabilization reserve mechanism. This indicates the carbon quota requirement capacity of the generator set;

[0201] The aforementioned restrictions on the behavior of power generators are as follows:

[0202]

[0203]

[0204] In the formula, This represents the lower limit of the selling price of generator set i in the carbon market within the bidding block. This represents the selling price of generator set i in the carbon market within the bidding block. This represents the upper limit of the selling price of generator set i in the carbon market within the bidding block. This represents the lower limit of the carbon market purchase price in the bidding block corresponding to generator set i. This represents the carbon market purchase price for generator set i in the bidding block. This represents the upper limit of the carbon market purchase price in the bidding block corresponding to generator set i. This indicates the lower limit of the price quoted for generator set i. This represents the bid price of generator set i in bidding block v. This indicates the upper limit of the price quoted for generator set i. express, express, This represents the selling price of generator set i in bidding block v-1;

[0205] The carbon quota accounting and scope constraints for the aforementioned power generators are as follows:

[0206]

[0207] In the formula, This represents the daily net surplus carbon allowance for generator set i. I represents the baseline value of carbon emission intensity for generator set i. i,t M0 represents the actual carbon emission intensity of generator unit i during time period t, and M0 represents the maximum carbon quota trading capacity of generator unit i. i This represents the binary variable indicating the carbon trading direction of generator set i, when u i When u = 1, it indicates that generator unit i is currently mainly selling carbon quotas; when u = 1, it indicates that generator unit i is mainly selling carbon quotas. i =0 indicates that generator set i is currently at the purchase position, a i b represents the coefficient of the carbon emission intensity constant term for generator set i. i This represents the linear coefficient of carbon emission intensity for generator set i.

[0208] Specifically, the aforementioned second constraint includes: power balance constraint, generator output constraint, total output power constraint, generator ramp capacity constraint, line transmission power constraint, node phase angle constraint, reference node constraint, carbon emission balance equation constraint, upper and lower limits constraint on carbon emission trading volume, and upper and lower limits constraint on total carbon emission capacity; among which, the aforementioned power balance constraint is:

[0209]

[0210] In the formula, This represents the total output power of the generator set. This represents the load demand of node n during time period t. It means that B n,m δ represents the susceptance between node n and node m. n,t δ represents the voltage phase angle of node n during time period t. m,t This represents the voltage phase angle of node m during time interval t. This represents the Lagrange multiplier associated with the aforementioned power balance constraints;

[0211] The output constraints of the above generator sets are:

[0212]

[0213] In the formula, This represents the minimum generating power of generator set i in bidding block v. This represents the maximum power output of generator set i in bidding block v. and All of these represent Lagrange multipliers related to the output constraints of the aforementioned generator sets;

[0214] The total output power constraint mentioned above is:

[0215]

[0216] In the formula, This represents the lower limit of the total output power of generator set i during time period t. This represents the upper limit of the total output power of generator set i during time period t. and Represents the dual variable;

[0217] The above generator ramp capacity constraint is:

[0218]

[0219] In the formula, This represents the upper limit of the power reduction rate of generator set i. This represents the output power of generator unit i in bidding block v during time period t+1. This represents the upper limit of the power increase rate of generator set i. Represents the dual variable. Represents the dual variable;

[0220] The above-mentioned line transmission power constraints are:

[0221]

[0222] In the formula, This represents the maximum transmission capacity of the line between node n and node m. Represents the dual variable;

[0223] The phase angle constraints of the above nodes are:

[0224]

[0225] In the formula, Represents the dual variable. Represents the dual variable;

[0226] The constraints of the above reference nodes are:

[0227]

[0228] In the formula, δ 1,t This represents the phase angle value of the reference node during time period t. Represents the dual variable;

[0229] The carbon emission balance equation constraints mentioned above are:

[0230]

[0231] In the formula, λ C This represents the Lagrange multiplier associated with the constraints of the carbon emission balance equation mentioned above;

[0232] The above-mentioned upper and lower limits for carbon emission trading are:

[0233]

[0234] In the formula, This represents the minimum amount of carbon allowance sold by generator set i. This represents the maximum amount of carbon allowance sold by generator set i. This represents the minimum carbon allowance that generator set i must purchase. This represents the maximum amount of carbon allowance that generator set i can purchase. and Represents the dual variable. and Represents the dual variable;

[0235] The upper and lower limits of the above total carbon emission capacity are:

[0236]

[0237] In the formula, and This represents the dual variable.

[0238] Step S103: Based on the above upper-level model and the current electric carbon coupling data, obtain the generator output that satisfies the above first constraint condition, and calculate the optimal electric carbon coupling clearing data corresponding to the above lower-level model based on the above generator output and the preset distribution mapping function.

[0239] Specifically, based on the current electric carbon coupling data, the upper-level model is solved under the first constraint condition to obtain the generator output that satisfies the first constraint condition. Then, it is substituted into the preset distribution mapping function. Since the preset distribution mapping function is used to quantify the impact of the upper-level decision on the lower level, the optimal electric carbon coupling clearing data corresponding to the lower-level model can be calculated after substitution.

[0240] Step S104: Based on the above optimal electro-carbon coupling clearing data and the current electro-carbon coupling data, solve the above upper-level model under the above first constraint condition to obtain the optimal generator set output, and schedule the generator set according to the optimal generator set output.

[0241] Specifically, after obtaining the optimal electric carbon coupling clearing data, it is combined with the current electric carbon coupling data to solve the upper-level model, thereby obtaining the optimal generator set output that satisfies the constraints. Finally, the optimal generator set output is used to achieve optimized scheduling of the generator sets.

[0242] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0243] like Figure 2 As shown, an embodiment of the present invention provides an electric carbon market coupled decision optimization device based on probability distribution mapping, comprising:

[0244] The system includes an electric carbon coupling data acquisition module, a model building module, an optimal electric carbon coupling clearing data calculation module, and a generator set scheduling module.

[0245] The aforementioned electro-carbon coupling data acquisition module is used to acquire the current electro-carbon coupling data;

[0246] The aforementioned model building module is used to construct an electricity-carbon price coupling decision model and corresponding constraints based on current electricity-carbon coupling data, with the objectives of maximizing generator profits, minimizing power dispatch costs, and optimizing the total value of supply and demand in carbon market quota trading. The electricity-carbon price coupling decision model includes an upper-level model with the objective of maximizing generator profits and a lower-level model with the objectives of minimizing power dispatch costs and optimizing the total value of supply and demand in carbon market quota trading. The constraints include a first constraint corresponding to the upper-level model and a second constraint corresponding to the lower-level model.

[0247] The above-mentioned optimal electric carbon coupling clearing data calculation module is used to obtain the generator output that satisfies the above-mentioned first constraint condition based on the above-mentioned upper-level model and the current electric carbon coupling data, and to calculate the optimal electric carbon coupling clearing data corresponding to the above-mentioned lower-level model based on the above-mentioned generator output and the preset distribution mapping function.

[0248] The generator set scheduling module is used to solve the upper-level model under the first constraint condition based on the optimal electric carbon coupling clearing data and the current electric carbon coupling data to obtain the optimal generator set output, and to schedule the generator sets according to the optimal generator set output.

[0249] It should be noted that the device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort. The above schematic diagram is merely an example of a probability distribution mapping-based electric carbon market coupled decision optimization device and does not constitute a limitation on a probability distribution mapping-based electric carbon market coupled decision optimization device. It may include more or fewer components than illustrated, or combine certain components, or use different components.

[0250] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0251] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-described method for coupled decision optimization of the electric carbon market based on probability distribution mapping in any embodiment of the present invention.

[0252] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the device.

[0253] The aforementioned terminal devices may be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. These devices may include, but are not limited to, processors and memory.

[0254] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the device, connecting various parts of the device via various interfaces and lines.

[0255] The aforementioned memory can be used to store the aforementioned computer programs and / or modules. The aforementioned processor implements various functions of the aforementioned device by running or executing the computer programs and / or modules stored in the aforementioned memory, and by calling data stored in the memory. The aforementioned memory may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0256] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0257] Another embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the above-described method for coupled decision optimization of the electric carbon market based on probability distribution mapping, as described in any embodiment of the present invention.

[0258] In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0259] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for coupling decision optimization of electricity and carbon markets based on probability distribution mapping, characterized in that, The method comprises the following steps: obtaining current electricity-carbon coupling data; building an electricity-carbon price coupling decision model and corresponding constraint conditions according to the current electricity-carbon coupling data, with the goal of maximizing the profits of power suppliers, minimizing the cost of power scheduling, and optimizing the total value of both sides of carbon market quota transaction supply and demand; wherein the electricity-carbon price coupling decision model comprises an upper model with the goal of maximizing the profits of power suppliers, and a lower model with the goal of minimizing the cost of power scheduling and optimizing the total value of both sides of carbon market quota transaction supply and demand; the constraint conditions comprise first constraint conditions corresponding to the upper model, and second constraint conditions corresponding to the lower model; obtaining generator unit output satisfying the first constraint conditions according to the upper model and the current electricity-carbon coupling data, and calculating the optimal electricity-carbon coupling settlement data corresponding to the lower model according to the generator unit output and a preset distribution mapping function; solving the upper model under the first constraint conditions according to the optimal electricity-carbon coupling settlement data and the current electricity-carbon coupling data, obtaining optimal generator unit output, and scheduling the generator unit according to the optimal generator unit output.

2. The method of claim 1, wherein, The construction of the preset distribution mapping function comprises: obtaining historical electricity-carbon coupling data in a historical period; repeating the distribution mapping function construction operation according to the historical electricity-carbon coupling data until the root mean square error of the current distribution mapping function is not greater than a preset threshold, and taking the current distribution mapping function as the preset distribution mapping function; wherein the distribution mapping function construction operation comprises: obtaining historical electricity-carbon coupling data of a current historical period and current generator unit optimization output; wherein the historical electricity-carbon coupling data at the initial time is the historical electricity-carbon coupling data corresponding to the initial historical period, and the generator unit optimization output at the initial time is a preset initial generator unit optimization output; fitting to obtain the current distribution mapping function according to the historical electricity-carbon coupling data of the current historical period; calculating the root mean square error of the current distribution mapping function according to the current generator unit optimization output, the current distribution mapping function, and the true electricity-carbon coupling settlement data in the historical electricity-carbon coupling data of the current historical period; in the case where the root mean square error of the current distribution mapping function is greater than the preset threshold, building an electricity-carbon price coupling decision model corresponding to the current historical period and corresponding current historical constraint conditions according to the historical electricity-carbon coupling data of the current historical period, with the goal of maximizing the profits of power suppliers, minimizing the cost of power scheduling, and optimizing the total value of both sides of carbon market quota transaction supply and demand; wherein the electricity-carbon price coupling decision model corresponding to the current historical period comprises a current historical upper model with the goal of maximizing the profits of power suppliers, and a current historical lower model with the goal of minimizing the cost of power scheduling and optimizing the total value of both sides of carbon market quota transaction supply and demand; the current historical constraint conditions comprise first current historical constraint conditions corresponding to the current historical upper model, and second current historical constraint conditions corresponding to the current historical lower model; According to the current generator set optimization output, the current history lower model is solved under the second current historical constraint condition, and current electric-carbon coupling clearing optimization data is obtained; According to the current electric-carbon coupling clearing optimization data and the historical electric-carbon coupling data of the current historical period, the electric-carbon price coupling decision model corresponding to the current historical period is solved under the current historical constraint condition, and the generator set optimization output corresponding to the next historical period is obtained.

3. The method of claim 2, wherein, The current distribution mapping function is fitted according to the historical electric-carbon coupling data of the current historical period, and the current distribution mapping function is fitted according to the historical electric-carbon coupling data of the current historical period, and the current distribution mapping function is fitted according to the historical electric-carbon coupling data of the current historical period. The historical electric-carbon clearing response parameter corresponding to the current historical period is extracted from the historical electric-carbon coupling data of the current historical period; The probability density function of the current historical period is calculated according to the historical electric-carbon coupling data of the current historical period; The joint probability density function of the current historical period is calculated according to the historical electric-carbon clearing response parameter corresponding to the current historical period; The distribution difference quantization index value of the current historical period is calculated according to the probability density function of the current historical period and the joint probability density function of the current historical period; The initial dynamic weight matrix coefficient corresponding to the current historical period is fitted and calculated according to the preset time base function, the preset space base function and the historical electric-carbon clearing response parameter corresponding to the current historical period; The dynamic weight matrix coefficient corresponding to the current historical period is calculated according to the initial dynamic weight matrix coefficient corresponding to the current historical period and the distribution difference quantization index value of the current historical period; The current distribution mapping function is constructed according to the dynamic weight matrix coefficient corresponding to the current historical period, the preset time base function and the preset space base function.

4. The method of claim 3, wherein, The current distribution mapping function is constructed according to the dynamic weight matrix coefficient corresponding to the current historical period, the preset time base function and the preset space base function, and the current distribution mapping function is constructed according to the dynamic weight matrix coefficient corresponding to the current historical period, the preset time base function and the preset space base function. The rated output of each generator set is obtained; The memory attenuation matrix is calculated according to the rated output of each generator set and the corresponding preset carbon emission parameter; The actual carbon emission intensity of each unit in the current historical period is extracted from the historical electric-carbon coupling data of the current historical period; The to-be-fitted distribution mapping function corresponding to the current historical period is constructed according to the memory attenuation matrix, the dynamic weight matrix coefficient corresponding to the current historical period, the preset time base function and the preset space base function, wherein the to-be-fitted distribution mapping function has a fractional order residual term; The to-be-fitted distribution mapping function corresponding to the current historical period is fitted according to the actual carbon emission intensity of each unit in the current historical period, and a plurality of weight coefficients of the distribution mapping function corresponding to the current historical period are determined. The current distribution mapping function is constructed according to the fractional order residual term, the plurality of weight coefficients, the memory attenuation matrix, the dynamic weight matrix coefficient corresponding to the current historical period, the preset random disturbance term, the preset time base function and the preset space base function.

5. The method of claim 4, wherein, The current distribution mapping function is constructed according to the fractional order residual term, a plurality of weight coefficients, a memory attenuation matrix, dynamic weight matrix coefficients corresponding to a current historical period, a preset random disturbance term, a preset time base function and a preset space base function, and the current distribution mapping function comprises: Derivation is performed on the fractional order residual term to obtain a derivative of the fractional order residual term; The fractional order residual term is calculated by using a numerical integration method and the derivative to obtain an approximate value of the fractional order residual term; A linear equation of the fractional order residual term is constructed according to the approximate value, the derivative, the fractional order residual term, a plurality of weight coefficients, a memory attenuation matrix, dynamic weight matrix coefficients corresponding to a current historical period, a preset random disturbance term, a preset time base function and a preset space base function; The linear equation is solved to obtain a solution of the fractional order residual term; The current distribution mapping function is constructed according to the solution of the fractional order residual term, a plurality of weight coefficients, a memory attenuation matrix, dynamic weight matrix coefficients corresponding to a current historical period, a preset random disturbance term, a preset time base function and a preset space base function.

6. The method of claim 5, wherein, The root mean square error of the current distribution mapping function is calculated according to the current generator set optimized output, the current distribution mapping function and the real electric-carbon coupling out-clear data in the historical electric-carbon coupling data of the current historical period, and the root mean square error of the current distribution mapping function comprises: The current electric-carbon coupling out-clear prediction data is calculated according to the current generator set optimized output and the current distribution mapping function; The root mean square error of the current distribution mapping function is calculated according to the current electric-carbon coupling out-clear prediction data and the real electric-carbon coupling out-clear data.

7. The method of claim 6, wherein, The lower layer model comprises a first lower layer model and a second lower layer model. The objective function of the upper layer model is: wherein represents the profit of the power producer g in the electricity market, represents the revenue of the power producer g in the carbon market, represents the cost of the power producer g in the carbon market; The objective function of the first lower layer model is: wherein represents the offer of the generator unit i in the bidding block v, represents the output power of the generator unit i in the bidding block v at time period t; The objective function of the second lower layer model is: wherein represents the buy price of the carbon market corresponding to the generator unit i in the bidding block, represents the amount of carbon allowances bought by the generator unit i, represents the sell price of the carbon market corresponding to the generator unit i in the bidding block, represents the amount of carbon allowances sold by the generator unit i.

8. A device for optimizing coupling decisions in an electricity-carbon market based on probability distribution mapping, characterized in that, It comprises: An electric-carbon coupling data acquisition module, a model construction module, an optimal electric-carbon coupling out-clear data calculation module and a generator set scheduling module; The electric-carbon coupling data acquisition module is configured to acquire current electric-carbon coupling data; The model construction module is configured to construct an electric-carbon price coupling decision model and corresponding constraint conditions according to the current electric-carbon coupling data, with the aim of maximizing the profits of power suppliers, minimizing the power scheduling cost and optimizing the total value of both sides of the carbon market quota transaction; wherein the electric-carbon price coupling decision model comprises an upper layer model aiming to maximize the profits of power suppliers and a lower layer model aiming to minimize the power scheduling cost and optimize the total value of both sides of the carbon market quota transaction; the constraint conditions comprise a first constraint condition corresponding to the upper layer model and a second constraint condition corresponding to the lower layer model; The optimal electric-carbon coupling out-clear data calculation module is configured to obtain a generator set output satisfying the first constraint condition according to the upper layer model and the current electric-carbon coupling data, and to calculate the optimal electric-carbon coupling out-clear data corresponding to the lower layer model according to the generator set output and a preset distribution mapping function. The generator set scheduling module is configured to solve the upper model under the first constraint condition according to the optimal electricity-carbon coupling out-clear data and current electricity-carbon coupling data, obtain optimal generator set output, and schedule the generator set according to the optimal generator set output.

9. A terminal device, comprising: A computer program product comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the computer program, when executed by the processor, implements the method of any one of claims 1 to 7.

10. A storage medium, characterized by The storage medium comprises a stored computer program, wherein the computer program, when executed, controls a device in which the storage medium is located to implement the method of any one of claims 1 to 7.

Images