Spot electric quantity global optimization method, system and device and storage medium

Abstract

The invention discloses a spot electric quantity global optimization method, system and device and a storage medium, and the method comprises the steps: decomposing the planned generating capacity, and generating an electric quantity combination curve; the electric quantity combination curve is input into a coal blending combustion model and a preset electricity price prediction model, a minimum-cost coal blending combustion scheme and an electricity price combination prediction curve are output, the coal blending combustion model is constructed based on a linear programming model, and the preset electricity price prediction model is constructed based on an ARIMA model and an LSTM network; and according to the electric quantity combination curve, the minimum cost coal blending combustion scheme and the electricity price combination prediction curve, an optimal benefit coal blending scheme is obtained through optimizing the objective function. According to the method, the coal blending combustion scheme with the minimum cost is output through the coal blending combustion model, the power generation cost is reduced, the accuracy of electricity price prediction is effectively improved through the ARIMA model and the LSTM network, and then the output result of the model is integrated through the optimization objective function, so that the economic benefits and market competitiveness of power generation enterprises are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of coal-fired power generation, and in particular to a global optimization method, system, device and storage medium for spot electricity. Background Art

[0002] In the power industry, efficient operation of power generation companies is crucial. Coal blending and electricity price forecasting are key to optimizing power generation costs and improving economic efficiency. Traditional coal blending relies primarily on experience, using relatively simple models or relying on manual experience to develop plans. Electricity price forecasting, which relies on a single model or simple statistical analysis, cannot effectively address the complex and volatile power market. This results in low electricity price forecast accuracy and cannot provide a reliable basis for power generation planning and cost control for power generation companies. Therefore, a new technical solution is urgently needed to obtain the most efficient coal blending plan to improve the economic efficiency and market competitiveness of power generation companies.

[0003] The above content is only used to assist in understanding the technical solution of the present invention and does not constitute an admission that the above content is prior art. Summary of the Invention The main purpose of the present invention is to provide a global optimization method, system, equipment and storage medium for spot electricity, aiming to solve the technical problem of how to obtain the most efficient coal blending solution, thereby improving the economic benefits and market competitiveness of power generation enterprises.

[0004] To achieve the above object, the present invention provides a global optimization method for spot electricity, which includes: Decompose the planned power generation of power generation enterprises and generate power combination curves; The electricity quantity combination curve is respectively input into the coal blending and combustion model and the preset electricity price prediction model, and the minimum cost coal blending and combustion scheme and the electricity price combination prediction curve are output. The coal blending and combustion model is constructed based on the linear programming model, and the preset electricity price prediction model is constructed based on the combination of the ARIMA model and the LSTM network; The optimal benefit coal blending scheme is obtained by searching the objective function according to the power combination curve, the minimum cost coal blending and combustion scheme and the electricity price combination prediction curve.

[0005] Optionally, decomposing the planned power generation of the power generation enterprise to generate a power combination curve includes: Calculate historical load data based on historical electricity data of coal-fired power plants; Based on the historical load data, the planned power generation of the power generation enterprise is decomposed using a preset decomposition curve to obtain a preset time period power sequence; An electric quantity combination curve is generated according to the electric quantity sequence of the preset time period.

[0006] Optionally, inputting the power combination curve into a coal blending and combustion model to output a minimum cost coal blending and combustion plan includes: Obtaining the maximum power generation capacity and coal type information of the coal-fired power plant; Calculating the load rate corresponding to the electricity quantity in each time period in the electricity quantity combination curve according to the maximum power generation capacity, and establishing the coal mill output constraint according to the load rate; According to the coal type information, coal calorific value constraint, sulfur content constraint, volatile matter constraint, blending ratio constraint and coal inventory constraint are respectively established; Based on the coal mill output constraint, the coal calorific value constraint, the sulfur content constraint, the volatile matter constraint, the blending ratio constraint and the coal inventory constraint, a coal blending and combustion model is constructed through the objective function of a linear programming model; The minimum cost coal blending and combustion plan is outputted through the coal blending and combustion model.

[0007] Optionally, the step of inputting the electricity quantity combination curve into a preset electricity price prediction model and outputting the electricity price combination prediction curve includes: Generate historical electricity quantity series and historical electricity price series based on historical electricity data of coal-fired power plants; Normalizing the historical electricity price sequence to obtain a normalized electricity price sequence; Determining the ARIMA model order and parameter estimation according to the normalized electricity price series; Establishing a linear regression equation based on the planned power generation sequence and the electricity price forecast sequence based on the ARIMA model order and parameters, and constructing an ARIMA model based on the linear regression equation; Training the LSTM initial network according to the historical electricity quantity sequence and the historical electricity price sequence, and using the trained LSTM initial network as the LSTM network; A preset electricity price prediction model is constructed based on the ARIMA model and the LSTM network.

[0008] Optionally, determining the ARIMA model order and parameter estimation according to the normalized electricity price series includes: Performing a stationary test on the normalized electricity price series; If it is not stable, performing order difference processing on the normalized electricity price series; generating an ACF graph and a PACF graph according to the processed normalized electricity price sequence; Determine the ARIMA model order according to the ACF graph and the PACF graph; Parameter estimation is obtained based on the model order and the processed normalized electricity price sequence by using a maximum likelihood estimation method or a least squares method.

[0009] Optionally, inputting the electricity quantity combination curve into a preset electricity price prediction model and outputting the electricity price combination prediction curve includes: Normalizing the electric quantity sequence of the preset time period corresponding to the electric quantity combination curve to obtain a normalized electric quantity sequence; Inputting the normalized electricity quantity sequence into the ARIMA model in the preset electricity price prediction model, and outputting the ARIMA predicted electricity price sequence; Inputting the normalized electricity quantity sequence and the ARIMA predicted electricity price sequence into the LSTM network in the preset electricity price prediction model, and outputting the LSTM predicted electricity price sequence; Calculating a weighted average prediction value based on the ARIMA predicted electricity price sequence and the LSTM predicted electricity price sequence; An electricity price combination prediction curve is generated according to the weighted average prediction value.

[0010] Optionally, the optimization objective function is:

[0011] Where, For the minimum cost coal blending scheme, is the number of coal types, is the number of time periods, is the unit cost of the i-th type of coal, is the blending ratio of the i-th coal in the t-th period, is the power generation in the tth period in the power combination curve, is the electricity price forecast value for the tth period in the electricity price combination forecast curve, is the minimum power generation in the tth period, is the maximum power generation in the tth period, is the inventory of type i coal.

[0012] In addition, to achieve the above-mentioned purpose, the present invention further proposes a global optimization system for spot electricity, which includes: The decomposition module is used to decompose the planned power generation of the power generation enterprise and generate a power combination curve; A prediction module is used to input the electricity quantity combination curve into a coal blending and combustion model and a preset electricity price prediction model, respectively, and output a minimum cost coal blending and combustion plan and an electricity price combination prediction curve, wherein the coal blending and combustion model is constructed based on a linear programming model, and the preset electricity price prediction model is constructed based on a combination of an ARIMA model and an LSTM network; The optimization module is used to obtain the optimal benefit coal blending plan through the optimization objective function according to the power combination curve, the minimum cost coal blending and combustion plan and the electricity price combination prediction curve.

[0013] In addition, to achieve the above-mentioned purpose, the present invention also proposes a global optimization device for spot electricity, which includes: a memory, a processor, and a global optimization program for spot electricity stored in the memory and runnable on the processor, and the global optimization program for spot electricity is configured to implement the steps of the global optimization method for spot electricity as described above.

[0014] In addition, to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a global optimization program for spot electricity is stored. When the global optimization program for spot electricity is executed by a processor, the steps of the global optimization method for spot electricity described above are implemented.

[0015] The present invention first decomposes the planned power generation of the power generation enterprise to generate a power combination curve, and then inputs the power combination curve into the coal blending and combustion model and the preset electricity price prediction model, outputting the minimum cost coal blending and combustion plan and the electricity price combination prediction curve. The coal blending and combustion model is constructed based on a linear programming model, and the preset electricity price prediction model is constructed based on an ARIMA model and an LSTM network combination. Then, the optimal benefit coal blending plan is obtained by optimizing the objective function based on the power combination curve, the minimum cost coal blending and combustion plan, and the electricity price combination prediction curve. Compared with the existing technology that uses relatively simple models or relies on manual experience to formulate plans and simple statistical analysis, it cannot effectively cope with the complex and changing situations in the power market, resulting in low accuracy of electricity price prediction and inability to provide a reliable basis for power generation planning and cost control of power generation enterprises. The present invention outputs the minimum cost coal blending and combustion plan through the coal blending and combustion model, reduces power generation costs, effectively improves the accuracy of electricity price prediction through the ARIMA model and the LSTM network, and then integrates the model output results through the optimization objective function, thereby improving the economic benefits and market competitiveness of power generation enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 1 is a schematic structural diagram of a global optimization device for spot electricity in a hardware operating environment according to an embodiment of the present invention; Figure 2 This is a flow chart of the first embodiment of the global optimization method for spot electricity quantity according to the present invention; Figure 3 This is a structural block diagram of the first embodiment of the global optimization system for spot electricity volume of the present invention.

[0017] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a global optimization device for spot electricity in the hardware operating environment involved in an embodiment of the present invention.

[0020] like Figure 1 As shown, the global optimization device for spot electricity consumption may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display and an input unit, such as a keyboard. Optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a wireless fidelity (Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also be a storage system independent of the processor 1001.

[0021] Those skilled in the art will understand that Figure 1 The structure shown in does not constitute a limitation on the global optimization device for spot electricity, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0022] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a global optimization program for spot electricity.

[0023] exist Figure 1 In the spot electricity global optimization device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the spot electricity global optimization device of the present invention can be set in the spot electricity global optimization device, and the spot electricity global optimization device calls the spot electricity global optimization program stored in the memory 1005 through the processor 1001, and executes the spot electricity global optimization method provided by the embodiment of the present invention.

[0024] The embodiment of the present invention provides a global optimization method for spot electricity. Figure 2 , Figure 22 is a flow chart of the first embodiment of the global optimization method for spot electricity quantity according to the present invention.

[0025] In this embodiment, the global optimization method for spot electricity includes the following steps: Step S10: Decomposing the planned power generation of the power generation enterprise to generate a power combination curve.

[0026] It is easy to understand that the execution subject of this embodiment can be a global optimization system for spot electricity with functions such as data processing, network communication and program running, or other computer equipment with similar functions, etc., and this embodiment is not limited to this.

[0027] Furthermore, the planned power generation of power generation companies is decomposed to generate a power combination curve. This processing method involves using historical power data of coal-fired power plants to calculate historical load data. Based on this historical load data, the power generation companies' planned power generation is decomposed using a preset decomposition curve to obtain a preset time period power sequence (including peak power sequence, flat power sequence, and valley power sequence). Finally, the power combination curve is generated based on the preset time period power sequence.

[0028] It should be noted that the preset decomposition curve may be a typical decomposition curve, such as the peak-flat-valley curve D5.

[0029] The processing method for calculating historical load data based on historical electricity data is to determine the generation time of historical electricity, the power generation of the whole day and the maximum output power of the generator set based on the historical electricity data, calculate the average power based on the generation time and the power generation of the whole day, and then calculate the historical load rate (i.e., historical load data) based on the average power and the maximum output power of the generator set.

[0030] It should also be understood that the planned power generation of the power generation enterprise needs to be broken down into a preset number of time periods. If the preset number of time periods is 96, then each time period is 15 minutes. Then, based on the historical load data, the 96 periods are divided into peak periods, flat periods and valley periods through the peak-flat-valley curve.

[0031] In this example, assume that the total contracted electricity consumption for the second day is 10,000 MWh. Peak-valley curve D5 is used for decomposition: Peak period (9:00 AM - 6:00 PM): 50% of the electricity consumption, or 5,000 MWh; Flat period (0:00 AM - 9:00 AM and 6:00 PM - 12:00 PM): 30% of the electricity consumption, or 3,000 MWh; Valley period (the rest of the time): 20% of the electricity consumption, or 2,000 MWh. The total electricity consumption is decomposed into 96 time periods (each 15 minutes long). The electricity data for each period (electricity consumption and time period) is then used to generate an electricity sequence, and a combined electricity curve is generated based on the electricity sequence.

[0032] Step S20: Input the electricity quantity combination curve into the coal blending and combustion model and the preset electricity price prediction model respectively, and output the minimum cost coal blending and combustion plan and the electricity price combination prediction curve. The coal blending and combustion model is constructed based on the linear programming model, and the preset electricity price prediction model is constructed based on the combination of the ARIMA model and the LSTM network.

[0033] Furthermore, the electricity combination curve is input into the coal blending and combustion model, and the processing method for outputting the minimum cost coal blending and combustion plan is to obtain the maximum power generation capacity and coal type information of the coal-fired power plant; calculate the load rate corresponding to the electricity in each period in the electricity combination curve according to the maximum power generation capacity, and construct the mill output constraint according to the load rate; construct the coal calorific value constraint, sulfur content constraint, volatile matter constraint, blending ratio constraint and coal inventory constraint according to the coal type information; construct the coal blending and combustion model based on the objective function of the linear programming model based on the mill output constraint, coal calorific value constraint, sulfur content constraint, volatile matter constraint, blending ratio constraint and coal inventory constraint; and output the minimum cost coal blending and combustion plan through the coal blending and combustion model.

[0034] The processing method for calculating the load rate corresponding to the electricity quantity in each time period within the electricity combination curve based on the maximum power generation capacity may be to calculate the average power corresponding to each time period based on the electricity quantity and duration of each time period, and to calculate the load rate corresponding to the electricity quantity in each time period based on the average power corresponding to each time period and the maximum power generation capacity.

[0035] Coal mill output constraints:

[0036] Where, is the output ratio (i.e. load rate) of the j-th coal mill in the t-th period, is the maximum output ratio of the j-th coal mill.

[0037] Based on the coal type information, the calorific value of coal, the sulfur content of coal, the type information of coal, the volatile matter of coal and the inventory of coal are determined. The calorific value constraint is constructed based on the calorific value of coal, the sulfur content constraint is constructed based on the sulfur content of coal, the blending ratio constraint is constructed based on the type information of coal, the volatile matter constraint is constructed based on the volatile matter of coal, and the coal inventory constraint is constructed based on the inventory of coal.

[0038] Heat generation limit:

[0039] Where, is the minimum value of heat output, is the maximum value of heat output, is the calorific value of the i-th coal, is the blending ratio of the i-th coal in the t-th time period, and n is the number of coal types.

[0040] Sulfur content constraints:

[0041] Where, is the sulfur content of the i-th coal, The upper limit of sulfur content.

[0042] Blending ratio constraints:

[0043] Volatile matter constraints:

[0044] Where, is the minimum value of volatile matter, is the maximum value of volatile matter, is the volatile matter of the i-th coal.

[0045] Coal inventory constraints:

[0046] Where, is the number of decomposed periods, is the power generation in the tth period, is the inventory of type i coal.

[0047] Objective function:

[0048] Where, For the minimum cost coal blending scheme, is the unit cost of the i-th type of coal.

[0049] In practice, the power generation curve, coal type information, and mill parameters are input into the coal blending and combustion model. This model is then solved using a linear programming solver (such as the linprog function in MATLAB or the scipy.optimize.linprog function in Python). Based on the objective function and constraints, the solver calculates the minimum-cost coal blending and combustion plan (including minimum cost, optimal blending ratios for each time period, and mill output ratios).

[0050] Furthermore, historical electricity quantity series and historical electricity price series are generated based on the historical electricity data of the coal-fired power plant; the historical electricity price series is normalized to obtain a normalized electricity price series; the ARIMA model order and parameter estimation of the AutoRegressive Integrated Moving Average (ARIMA) model are determined based on the normalized electricity price series; a linear regression equation is established based on the ARIMA model order and parameters according to the planned power generation series and the electricity price forecast series, and an ARIMA model is constructed based on the linear regression equation; a Long Short-Term Memory (LSTM) initial network is trained based on the historical electricity quantity series and the historical electricity price series, and the trained LSTM initial network is used as the LSTM network; and a preset electricity price forecast model is constructed based on the ARIMA model and the LSTM network.

[0051] In this embodiment, the historical electricity price sequence is , the historical electricity price series corresponds to the historical electricity quantity series: , these data should have the same time span and time interval, for example, data within a certain period recorded in days.

[0052] It is also important to check whether there are missing values ​​or outliers in the data. For missing values, methods such as mean filling and linear interpolation can be used to handle them. For outliers, whether to retain or correct them should be determined based on the actual situation.

[0053] The historical electricity price series are And the historical electricity price series corresponding to the historical electricity series is Perform normalization processing, the normalization formula is:

[0054] Where, is the normalized electricity price series, is the normalized electrical quantity series.

[0055] Furthermore, the processing method for determining the ARIMA model order and parameter estimation based on the normalized electricity price series is to perform a stationarity test on the normalized electricity price series; if it is not stationary, perform order difference processing on the normalized electricity price series; generate an autocorrelation function graph (Autocorrelation FunctionACF) and a partial autocorrelation function graph (Partial Autocorrelation FunctionPACF) based on the processed normalized electricity price series; determine the ARIMA model order based on the ACF graph and the PACF graph; and obtain parameter estimates based on the model order using the maximum likelihood estimation method or the least squares method according to the processed normalized electricity price series.

[0056] In the specific implementation, the normalized electricity price series is tested for stability; if it is not stable, the normalized electricity price series is processed by d-order difference to obtain the processed normalized electricity price series (i.e., the stable series). ) and plot its ACF and PACF. The ACF reflects the correlation between the series and its own lagged values, while the PACF reflects the correlation between the series and a specific lag value after removing the influence of intermediate lags. Determine the order (p, d, q) of the ARIMA model based on the characteristics of the ACF and PACF plots. If the ACF is trailing and the PACF is truncated after order p, the autoregressive order is p; if the PACF is trailing and the ACF is truncated after order q, the moving average order is q; d is the order of differencing used to stationary the data. For example, if the ACF approaches zero rapidly after lag 1 and the PACF approaches zero after lag 2, and first-order differencing was used to stationary the data, the model order is likely ARIMA(2,1,1).

[0057] Using stationary series , use appropriate parameter estimation methods, such as maximum likelihood estimation method, least squares method, etc., to estimate the parameter estimates in the ARIMA(p,d,q) model (including autoregressive coefficients φ1,φ2,...,φp and moving average coefficients θ1,θ2,...,θq).

[0058] In this embodiment, a linear regression equation is established based on the ARIMA model order and parameters according to the planned power generation sequence and the electricity price forecast sequence. The linear regression equation is:

[0059] Where, is the normalized planned power generation sequence (i.e., normalized power sequence), is the normalized electricity price forecast sequence (i.e., ARIMA forecast electricity price sequence), 、 、 are parameters estimated from historical data.

[0060] In this embodiment, the historical electricity quantity sequence and the historical electricity price sequence are normalized to obtain the normalized historical electricity quantity sequence and the normalized electricity price sequence. Input samples are constructed based on the normalized historical electricity quantity sequence and the normalized electricity price sequence. The samples are used to train the LSTM initial network, and the trained LSTM initial network is used as the LSTM network.

[0061] Furthermore, the electricity quantity combination curve is input into the preset electricity price prediction model, and the processing method of outputting the electricity price combination prediction curve is to normalize the electricity quantity sequence of the preset period corresponding to the electricity quantity combination curve to obtain a normalized electricity quantity sequence; the normalized electricity quantity sequence is input into the ARIMA model in the preset electricity price prediction model, and the ARIMA predicted electricity price sequence is output; the normalized electricity quantity sequence and the ARIMA predicted electricity price sequence are input into the LSTM network in the preset electricity price prediction model, and the LSTM predicted electricity price sequence is output. ; Calculate the weighted average forecast value based on the ARIMA forecast electricity price series and the LSTM forecast electricity price series; Generate the electricity price combination forecast curve based on the weighted average forecast value .

[0062]

[0063] Where, is the weight of the ARIMA model, is the weight of the LSTM network.

[0064] Step S30: Obtaining the most efficient coal blending scheme by searching the objective function according to the power combination curve, the minimum cost coal blending scheme and the electricity price combination prediction curve.

[0065] The optimization objective function is:

[0066] Where, For the minimum cost coal blending scheme, is the number of coal types, is the number of time periods, is the unit cost of the i-th type of coal, is the blending ratio of the i-th coal in the t-th period, is the power generation in the tth period in the power combination curve, is the electricity price forecast value for the tth period in the electricity price combination forecast curve, is the minimum power generation in the tth period, is the maximum power generation in the tth period, is the inventory of type i coal.

[0067] In this embodiment, the planned power generation of a power generation enterprise is first decomposed to generate a combined power curve. The combined power curve is then input into a coal blending and combustion model and a preset electricity price forecasting model, respectively, to output a minimum-cost coal blending and combustion plan and a combined electricity price forecasting curve. The coal blending and combustion model is constructed based on a linear programming model, while the preset electricity price forecasting model is constructed based on a combination of an ARIMA model and an LSTM network. The optimal benefit coal blending plan is then obtained by optimizing the objective function based on the combined power curve, the minimum-cost coal blending and combustion plan, and the combined electricity price forecasting curve. Compared to existing technologies that use simpler models or rely on manual experience to formulate plans and simple statistical analysis, which cannot effectively address the complex and volatile conditions in the power market, resulting in low accuracy in electricity price forecasts and an inability to provide a reliable basis for power generation planning and cost control at power generation enterprises, this embodiment uses the coal blending and combustion model to output a minimum-cost coal blending and combustion plan, thereby reducing power generation costs. The ARIMA model and LSTM network effectively improve the accuracy of electricity price forecasts. The model output results are then integrated through optimizing the objective function, thereby improving the economic benefits and market competitiveness of power generation enterprises.

[0068] Reference Figure 3 , Figure 3 This is a structural block diagram of the first embodiment of the global optimization system for spot electricity volume of the present invention.

[0069] like Figure 3 As shown, the global optimization system for spot electricity provided in the embodiment of the present invention includes: Decomposition module 3001, used to decompose the planned power generation of the power generation enterprise and generate a power combination curve; Prediction module 3002, configured to input the electricity quantity combination curve into a coal blending and combustion model and a preset electricity price prediction model, respectively, and output a minimum cost coal blending and combustion plan and an electricity price combination prediction curve, wherein the coal blending and combustion model is constructed based on a linear programming model, and the preset electricity price prediction model is constructed based on a combination of an ARIMA model and an LSTM network; The optimization module 3003 is used to obtain the most efficient coal blending scheme by optimizing the objective function according to the power combination curve, the minimum cost coal blending scheme and the electricity price combination prediction curve.

[0070] Other embodiments or specific implementations of the global optimization system for spot electricity of the present invention can refer to the above-mentioned method embodiments and will not be described in detail here.

[0071] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or system comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or system. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or system comprising the element.

[0072] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0073] Through the description of the above embodiments, those skilled in the art will clearly understand that the above-mentioned embodiments and methods can be implemented by means of software plus the necessary general-purpose hardware platform. Of course, hardware can also be used, but in many cases the former is a more preferred embodiment. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory / random access memory, magnetic disk, or optical disk) and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0074] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A global optimization method for spot electricity, characterized by: The global optimization method for spot electricity volume includes the following steps: Decompose the planned power generation of power generation enterprises and generate power combination curves; The electricity quantity combination curve is respectively input into the coal blending and combustion model and the preset electricity price prediction model, and the minimum cost coal blending and combustion scheme and the electricity price combination prediction curve are output. The coal blending and combustion model is constructed based on the linear programming model, and the preset electricity price prediction model is constructed based on the combination of the ARIMA model and the LSTM network; The optimal benefit coal blending scheme is obtained by searching the objective function according to the power combination curve, the minimum cost coal blending and combustion scheme and the electricity price combination prediction curve.

2. The method according to claim 1, wherein The method of decomposing the planned power generation of the power generation enterprise and generating a power combination curve includes: Calculate historical load data based on historical electricity data of coal-fired power plants; Based on the historical load data, the planned power generation of the power generation enterprise is decomposed using a preset decomposition curve to obtain a preset time period power sequence; An electric quantity combination curve is generated according to the electric quantity sequence of the preset time period.

3. The method according to claim 2, wherein The step of inputting the power combination curve into the coal blending and combustion model and outputting the minimum cost coal blending and combustion plan includes: Obtaining the maximum power generation capacity and coal type information of the coal-fired power plant; Calculating the load rate corresponding to the electricity quantity in each time period in the electricity quantity combination curve according to the maximum power generation capacity, and establishing the coal mill output constraint according to the load rate; According to the coal type information, coal calorific value constraint, sulfur content constraint, volatile matter constraint, blending ratio constraint and coal inventory constraint are respectively established; Based on the coal mill output constraint, the coal calorific value constraint, the sulfur content constraint, the volatile matter constraint, the blending ratio constraint and the coal inventory constraint, a coal blending and combustion model is constructed through the objective function of a linear programming model; The minimum cost coal blending and combustion plan is outputted through the coal blending and combustion model.

4. The method according to claim 2, wherein The step of inputting the electricity quantity combination curve into a preset electricity price prediction model and outputting the electricity price combination prediction curve includes: Generate historical electricity quantity series and historical electricity price series based on historical electricity data of coal-fired power plants; Normalizing the historical electricity price sequence to obtain a normalized electricity price sequence; Determining the ARIMA model order and parameter estimation according to the normalized electricity price series; Establishing a linear regression equation based on the planned power generation sequence and the electricity price forecast sequence based on the ARIMA model order and parameters, and constructing an ARIMA model based on the linear regression equation; Training the LSTM initial network according to the historical electricity quantity sequence and the historical electricity price sequence, and using the trained LSTM initial network as the LSTM network; A preset electricity price prediction model is constructed based on the ARIMA model and the LSTM network.

5. The method according to claim 4, wherein Determining the ARIMA model order and parameter estimation according to the normalized electricity price sequence includes: Performing a stationary test on the normalized electricity price series; If it is not stable, performing order difference processing on the normalized electricity price series; generating an ACF graph and a PACF graph according to the processed normalized electricity price sequence; Determine the ARIMA model order according to the ACF graph and the PACF graph; Parameter estimation is obtained based on the model order and the processed normalized electricity price sequence by using a maximum likelihood estimation method or a least squares method.

6. The method according to claim 2, wherein The step of inputting the electricity quantity combination curve into a preset electricity price prediction model and outputting the electricity price combination prediction curve comprises: Normalizing the electric quantity sequence of the preset time period corresponding to the electric quantity combination curve to obtain a normalized electric quantity sequence; Inputting the normalized electricity quantity sequence into the ARIMA model in the preset electricity price prediction model, and outputting the ARIMA predicted electricity price sequence; Inputting the normalized electricity quantity sequence and the ARIMA predicted electricity price sequence into the LSTM network in the preset electricity price prediction model, and outputting the LSTM predicted electricity price sequence; Calculating a weighted average prediction value based on the ARIMA predicted electricity price sequence and the LSTM predicted electricity price sequence; An electricity price combination prediction curve is generated according to the weighted average prediction value.

7. The method according to any one of claims 1 to 6, wherein: The optimization objective function is: Where, For the minimum cost coal blending scheme, is the number of coal types, is the number of time periods, is the unit cost of the i-th type of coal, is the blending ratio of the i-th coal in the t-th period, is the power generation in the tth period in the power combination curve, is the electricity price forecast value for the tth period in the electricity price combination forecast curve, is the minimum power generation in the tth period, is the maximum power generation in the tth period, is the inventory of type i coal.

8. A global optimization system for spot electricity, characterized by: The spot electricity global optimization system includes: The decomposition module is used to decompose the planned power generation of the power generation enterprise and generate a power combination curve; A prediction module is used to input the electricity quantity combination curve into a coal blending and combustion model and a preset electricity price prediction model, respectively, and output a minimum cost coal blending and combustion plan and an electricity price combination prediction curve, wherein the coal blending and combustion model is constructed based on a linear programming model, and the preset electricity price prediction model is constructed based on a combination of an ARIMA model and an LSTM network; The optimization module is used to obtain the optimal benefit coal blending plan through the optimization objective function according to the power combination curve, the minimum cost coal blending and combustion plan and the electricity price combination prediction curve.

9. A global optimization device for spot electricity, characterized in that: The device includes: a memory, a processor, and a spot electricity global optimization program stored in the memory and executable on the processor, wherein the spot electricity global optimization program is configured to implement the steps of the spot electricity global optimization method according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores a global optimization program for spot electricity. When the global optimization program for spot electricity is executed by the processor, the steps of the global optimization method for spot electricity as described in any one of claims 1 to 7 are implemented.

Images