Medium-and-long-term price prediction method based on long-period simulation of electricity market

Abstract

The invention relates to the technical field of electricity price prediction, in particular to an electricity market long-period simulation-based medium and long-term price prediction method, which comprises the following steps of S1, market basic data acquisition: acquiring whole market basic data information; s2, unit commitment model construction and day-ahead simulation: constructing a unit commitment model considering power grid security constraints; s3, economic dispatching model construction: constructing an economic dispatching model considering power grid security constraints; s4, node marginal electricity price and load flow calculation: node marginal electricity price calculation and load flow calculation are carried out; s5, uncertainty factor evaluation: the uncertainty factors existing in the electricity market are evaluated; s6, analyzing technical performance and economic indexes: analyzing statistics of various technical performance and economic indexes of the system; according to the method, the stability and the reliability of an enterprise under an emergency situation are enhanced, so that the enterprise can better adapt to market changes, and the scientificity and the accuracy of decision making are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electricity price prediction, in particular to a long-term price prediction method based on long-period simulation of a power market. BACKGROUND

[0002] With the rapid development of the power industry and the increasing market competition, the power spot market gradually becomes the core component of the power market system. In this complex and dynamic market environment, accurate prediction of power price changes is of great significance for market participants to develop effective trading strategies, reduce risks, and optimize resource allocation.

[0003] Long-term price prediction can be used to guide long-term trading and provide price reference, thus playing an important role in strategic planning, risk management, investment planning, etc. However, traditional price prediction mainly includes time series analysis and statistical models. However, the power market has complex interactions and influencing factors, and does not have the ability to handle complex nonlinear relationships. In order to solve these problems, constructing optimized SCUC and SCED models for market simulation has become an effective means, which provides accurate and reliable decision support for power market participants, and is expected to optimize resource allocation, reduce transaction risks, and promote the sustainable development of the power industry. SUMMARY

[0004] The present application provides a long-term price prediction method based on long-period simulation of a power market.

[0005] The long-term price prediction method based on long-period simulation of a power market comprises the following steps: S1, market basic data acquisition: acquiring full market basic data information, including power supply side data, load side data and fuel data; S2, unit commitment model construction and day-ahead simulation: based on the full market basic data information, considering various day-ahead predictions and maintenance arrangements, constructing a unit commitment model considering power grid safety constraints, and performing day-ahead simulation; S3, economic dispatching model construction: based on the full market basic data information, considering start-up arrangements and real-time market source network load data, constructing an economic dispatching model considering power grid safety constraints; S4, node marginal price and power flow calculation: based on the constructed unit commitment model considering power grid safety constraints and economic dispatching model considering power grid safety constraints, performing node marginal price calculation and power flow calculation; S5, uncertainty factor evaluation: based on the node marginal price calculation result, evaluating the uncertainty factors existing in the power market; S6, technical performance and economic index analysis: based on the results of the uncertainty factor evaluation, the statistics of various technical performances and economic indexes of the system are analyzed.

[0006] Optionally, the market basic data acquisition in S1 comprises: S11, power supply side data acquisition: through constructing a photovoltaic power generation output simulation model, based on the sunshine intensity and installed capacity, the historical year 8760-hour photovoltaic power generation curve is generated, and according to the installed capacity prediction value of the target planning year, the curve is adjusted in proportion to obtain the photovoltaic power generation output data of the planning year; S12, load side data acquisition: based on historical load data, the interpolation method is used to calculate the daily peak load in a year, and the actual change is simulated through random noise to generate 8760-hour load data in a historical year, and the load curve meeting the demand of the planning year is obtained according to the peak load and total electricity demand prediction value of the planning year.

[0007] Optionally, the power supply side data acquisition in S11 comprises: S111, constructing daily photovoltaic power generation characteristic curve: studying the sunshine intensity and sunshine hours of the local photovoltaic power station, assuming that the power generation capacity of the photovoltaic power station is m (ten thousand kilowatts), according to the sunshine intensity and sunshine hours, the daily photovoltaic power generation characteristic curve of the region is constructed, which is represented as: i (i=1, 2,...24); S112, constructing monthly photovoltaic maximum power generation output curve: according to the change characteristics of the light intensity and light time in four seasons (long light time in summer, short light time in winter), the monthly photovoltaic maximum power generation output curve is constructed, which is represented as: j (j=1, 2,...12); S113, constructing historical year solar power generation year time sequence output curve: combining the daily photovoltaic power generation characteristic curve and the monthly photovoltaic maximum power generation output curve, the 8760-hour power generation output curve of the historical year photovoltaic power station is constructed; S114, target planning year photovoltaic power generation output curve prediction: the target planning year photovoltaic power station installed capacity prediction value is set as n (ten thousand kilowatts), assuming that the light characteristics of the region do not change until the target planning year, according to the ratio of the power station installed capacity of the target planning year to the power generation installed capacity of the power station, the historical year 8760-hour power generation output curve is adjusted to obtain the target planning year photovoltaic power station 8760-hour power generation output curve, which is represented as: ; Wherein, P t,plan.solar (ten thousand kilowatts) is the photovoltaic power at time t of the target planning year, P t,hist.solarM (million kilowatt) is the photovoltaic power of the historical year t, m (million kilwatt) is the installed capacity of the solar power station, and n (million kilwatt) is the predicted value of the installed capacity of the solar power station in the target planning year.

[0008] Optionally, the load side data acquisition in S12 comprises: S121, constructing the annual load curve of the planning year: obtaining the monthly peak load data M of a historical year of a certain region or province j (j=1, 2...12) and the daily (winter and summer) 24-hour load data W of the certain region or province n and S n (n=1, 2, 3...24); S122, calculating the annual daily load peak data: calculating the annual daily load peak data Di (i=1, 2...365) by using the interpolation method; S123, simulating the actual load curve by adding noise: adding noise to the annual daily load peak data Di by using the random number method to obtain 365 daily load curves close to the actual ones, thereby forming the 8760-hour load data of the historical year; S124, calculating the 8760-hour load data of the historical year: assuming that the daily 24-hour load data of the summer (May-September) and the winter (October-April) are S n and W n , extracting the maximum load demand in the daily 24 hours as the daily load peak H i , according to and solving to obtain new daily 24-hour load data in the winter and the summer, thereby obtaining the 24-hour load data of 365 days, i.e., the 8760-hour load data H k (k=1, 2...8760) of the historical year; S125, predicting the load curve of the target planning year: assuming that the load peak of the historical year is D max , the predicted value of the load peak of the planning year is D , and the 8760-hour load data H of the planning year can be obtained by proportional adjustment; S126, adjusting the load curve to match the total electricity demand of the planning year: knowing the total electricity consumption C of the historical year and the predicted value C' of the electricity demand of the planning year, keeping the peak load of the planning year unchanged, adjusting the 8760-hour load curve according to the deviation degree, so that the total electricity demand of the annual continuous load curve meets the requirement of the predicted value of the target planning year electricity demand, and calculating the 8760-hour load curve of the planning year.

[0009] Optionally, the objective function of the unit commitment model considering the power grid safety constraint is represented as: ; where F[p i (t), I i (t)] is the total generation cost of the system, p i (t) and I i (t) are decision variables, p i (t) is the output of unit i at time period t, I i (t) is the on-off state of unit i at time period t, I i (t) = 1 indicates that the unit is in operation, I i (t) = 0 indicates that the unit is in shutdown, C i [p i (t)] is the generation operation cost of unit i at time period t, S i (t) is the start-up cost of unit i at time period t, M is the number of units, and T is the total number of time periods covered by the unit commitment problem. The constraint conditions of the unit commitment model considering grid security constraints include: System power balance constraint: ; where P d (t) is the total load of the system at time period t. Maximum and minimum output constraints of generators: ; where P imin and P imax are the minimum and maximum outputs of unit i, respectively. Unit spinning reserve capacity constraint: ; where S D (t) is the total reserve capacity demand of the system at time period t. Ramp rate limit constraint: ; where DR i and DR i are the allowed up and down outputs of unit i at each time period. Minimum operation and shutdown duration constraints of units: ; ; where T and T are the minimum operation and shutdown times of unit i, and T and T are the continuous on and off times of unit i before time period t. Renewable energy generation constraint: ; where P h is the output of renewable energy at time period t.The theoretical output of the renewable energy unit h; Power grid transmission safety constraint: ; Wherein, GSF l-i , GSF l-h , GSF l-k are the power generation transfer factors of the thermal power unit i, the renewable energy unit h and the load k to the transmission equipment l, respectively, F l,min and F l,max are the power flow and the upper and lower limits of the line i-j, respectively.

[0010] Optionally, the objective function of the economic dispatching model considering the power grid safety constraint is represented as: Wherein, NG is the number of system conventional units, p i and p h are the active power of the conventional unit i and the renewable energy unit h, respectively, C i (p i ) and C h (p h ) are the operation costs of the units i and h; The constraint conditions of the economic dispatching model considering the power grid safety constraint include: System load balance constraint: ; Wherein, D k is the bus load, and NLD is the number of load buses; Generator maximum and minimum output constraint: ; Wherein, pimin and pimax are the upper and lower limits of the output power of the generator unit i, respectively; Ramp rate limit constraint: ; Wherein, pi is the maximum value of the load that can be added or reduced by the unit i per period.

[0011] Optionally, the node marginal price in S4 and the power flow calculation include: S41, processing the output power of the non-pricing unit: for the non-pricing unit, the upper and lower limit constraints of the unit output in the economic dispatching model considering the power grid safety constraint are replaced by the fixed output constraint, which is represented as: ; Wherein, is the winning output of the unit i in period t in the day-ahead market SCED calculation result; S42, calculate the node marginal price: after the output of the non-pricing unit is fixed, the economic dispatching model considering the power grid safety constraint in the day-ahead market is recalculated to obtain the Lagrange multiplier of each period, and the node price of the node k in the period t is calculated, which is expressed as: ; wherein, is the Lagrange multiplier of the system load balance constraint in the period t, and are the Lagrange multipliers of the maximum forward and reverse flow constraints of the branch l respectively, and when the branch flow is out of limit, the Lagrange multiplier is the network flow constraint relaxation penalty factor, and are the Lagrange multipliers of the maximum forward and reverse flow constraints of the section s respectively, and when the section flow is out of limit, the Lagrange multiplier is the network flow constraint relaxation penalty factor, is the generator output power transfer distribution factor of the node k to the section s.

[0012] Optionally, the uncertainty factors in S5 include macro uncertainty factors and micro uncertainty factors.

[0013] Advantages of the present application: The present application effectively improves the quantification ability of coping with uncertainty through the target optimization model, solves the problem of long-term electricity price prediction, and through multi-source data fusion and comprehensive consideration of various dynamic factors related to electricity price, combined with multi-modal information, the obtained prediction result is more accurate, which significantly improves the accuracy of electricity price prediction of power enterprises, and at the same time, the method enhances the stability and reliability of enterprises in emergency conditions, so that it can better adapt to market changes and improve the scientificity and accuracy of decision-making.

[0014] The present application comprehensively integrates market basic data through the target optimization model, including the power supply side, the load side and the fuel side, and constructs the corresponding calculation model, so that it is more consistent with the complex real environment, improves the flexibility of power grid operation, enhances the ability of power system to cope with the fluctuation of renewable energy generation, improves the adaptability and reliability of power grid, and at the same time, the method fills the research gap of long-term electricity price prediction in the spot market, provides a solid theoretical basis and practical guidance for the industry, and has important practical application value and broad market prospect. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to make the technical solutions in the present application or prior art clearer, the accompanying drawings needed in the embodiments or prior art description will be briefly introduced below. Obviously, the accompanying drawings in the following description only aim at the present application, and other accompanying drawings can be obtained by those skilled in the art without any creative effort.

[0016] Figure 1 A flowchart of the prediction method of the embodiment of the present application is shown in the figure. Figure 2 A schematic diagram of the annual 8760-hour load curve of the embodiment of the present application is shown in the figure. Figure 3 A schematic diagram of the annual 8760-hour system marginal price of the embodiment of the present application is shown in the figure. Figure 4 A schematic diagram of the wind power time series of the embodiment of the present application is shown in the figure. Figure 5 A schematic diagram of the photovoltaic time series of the embodiment of the present application is shown in the figure. Figure 6 A schematic diagram of the comparison between the actual and simulated electricity prices in April of the embodiment of the present application is shown in the figure. Figure 7 A schematic diagram of the comparison between the actual and simulated electricity prices in May of the embodiment of the present application is shown in the figure. Figure 8 A schematic diagram of the comparison between the actual and simulated electricity prices in June of the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0017] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, in order to make the embodiments more detailed, the following embodiments are the best, preferred embodiments, and other alternative ways can also be used by those skilled in the art to implement them; and the accompanying drawings are only used to more specifically describe the embodiments, and are not intended to specifically limit the present application.

[0018] It should be noted that, in the specification, "one embodiment", "embodiment", "exemplary embodiment", "some embodiments" and the like indicate that the described embodiments can include specific features, structures or characteristics, but not necessarily every embodiment includes the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in combination with an embodiment, it should be within the knowledge of those skilled in the art to realize this feature, structure or characteristic in combination with other embodiments (whether or not explicitly described).

[0019] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0020] like Figures 1-8 As shown, the medium- to long-term price forecasting method based on long-term electricity market simulation includes the following steps: Step S1: Obtain basic market data information, including power supply data, load data, and fuel measurement data.

[0021] Obtaining power source data requires constructing a power generation output curve simulation model, studying the solar irradiance and sunshine hours characteristics of the local photovoltaic power station, setting the power generation capacity of the photovoltaic power station as m (ten thousand kilowatts), and constructing a typical daily photovoltaic power generation characteristic curve S for the region. i (i=1,2,...24); Secondly, based on the seasonal variations in photovoltaic irradiance and sunshine duration in the region, such as long sunshine duration in summer and short sunshine duration in winter, a monthly maximum photovoltaic power generation curve S is constructed. j (j=1,2,...12); then, based on the typical daily photovoltaic power generation characteristic curve and the monthly maximum photovoltaic power generation output curve, the historical annual time-series power generation output curve of the solar power generation is constructed, that is, the 8760-hour photovoltaic power generation output curve of the power station; finally, the predicted value of the photovoltaic power station installed capacity in the target planning year is set as n (ten thousand kilowatts). Under the assumption that the regional irradiance characteristics will not change significantly from the target planning year, the historical 8760-hour power generation output curve can be adjusted according to the ratio of the power station installed capacity in the target planning year to the power generation installed capacity of the power station, to obtain the 8760-hour power generation output curve of the photovoltaic power station in the target planning year, that is: ; In the formula, P t,plan.solar (10,000 kilowatts) represents the photovoltaic power generation capacity at time t in the target year; P t,hist.solar (MW) represents the photovoltaic power generation at time t in historical year; m (MW) represents the installed capacity of the solar power plant; n (MW) represents the predicted installed capacity of the solar power plant in the target planning year.

[0022] Obtaining load-side data requires constructing a load curve for the entire planning year to obtain historical monthly peak load data M for a specific region or province. j (j=1,2...12); Typical daily (winter and summer) 24-hour load data W for a certain region or province. nand S n (n=1,2,3...24); based on the known 12 months of monthly load peak data M j , the daily load peak data D of the whole year is obtained by interpolation i (i=1,2...365); the difference value of the interpolation method can be expressed as: ; Assuming that the load peak D1 of the first day of the first month is M1, then: ; ; The load peak D32 of the first day of the second month is M2, then: ; ; According to this process, the daily load peak data D of the whole year can be obtained i (i=1,2...365).

[0023] For the obtained daily load peak data D of the whole year i , and using the way of random number to add noise to it, 365 daily load curves close to the actual are obtained, that is, the historical year 8760 load data, the function of adding noise is: ; Where m is the noise multiple.

[0024] Assuming that the daily 24-hour load data in summer (May-September) and winter (October-April) is S n and W n , the maximum load demand in 24 hours is extracted as the load peak H i of the typical day; according to and , the new daily 24-hour load data in winter and summer is obtained, and thus the 365-day 24-hour load data is obtained, that is, the historical year 8760-hour load data H k (k=1,2...8760), wherein H k is: ; The historical year load peak is set as D max , and the load peak prediction value of the planning year is set as , then the new 8760-hour load data can be obtained by proportional adjustment, that is: ; The known historical year total power consumption C, the power demand forecast value C' of the planning year; keep the planning year peak load unchanged, adjust the 8760-hour load curve according to the deviation degree, so that the total power demand of the annual continuous load curve meets the requirement of the target planning year power demand forecast value. The deviation degree formula is: ; The planning year time sequence load curve formula is as follows, so as to meet the planning year total power demand forecast value: ; ; At this time, the 8760-hour load curve data of the planning year can be calculated.

[0025] According to the flow and the annual 8760-hour load diagram constructed according to the input peak load and power consumption, as shown in Figure 2 .

[0026] Step S2, based on the full market basic data, while considering various day-ahead forecasts and maintenance arrangements, a unit commitment model considering power grid safety constraints is constructed, and day-ahead simulation is performed.

[0027] The objective function of the unit commitment model considering power grid safety constraints is: ; In the formula, F[p i (t),I i (t)] is the total power generation cost of the system; p i (t) and I i (t) are decision variables, p i (t) is the output of unit i at t period, I i (t) is the start state of unit i at t period, I i (t)=1 indicates that the unit is in operation, I i (t)=0 indicates that the unit is in shutdown state, C i [p i (t)] is the power generation operation cost of unit i at t period, S i (t) is the start cost of unit i at t period; M is the number of units, and T is the total period number covered by the unit commitment problem.

[0028] In addition, the model also needs to meet other constraint conditions: 1) System power balance constraint: ; In the formula, P d (t) is the total load of the system at t period.

[0029] 2) Generator maximum and minimum power output constraints: ; where P imin and P imax are the minimum and maximum power output of unit i, respectively.

[0030] 3) Unit spinning reserve capacity constraints: ; where S D (t) is the total spinning reserve capacity requirement of the system at time t.

[0031] 4) Ramp rate limit constraints: ; where DR i and DR i are the allowed up and down power output of unit i at each time period, respectively.

[0032] 5) Minimum on- and off- duration constraints: For conventional thermal units, once off, it must be on for a certain period of time (minimum off-time) before it can be off again; conversely, once on, it must be off for a certain period of time (minimum on-time) before it can be on again. The minimum on-time and off-time constraints are as follows: ; ; where T and T are the minimum on- and off- time of unit i, respectively, and T and T are the on- and off- time of unit i before time t, respectively.

[0033] 6) Renewable energy generation constraints: ; where P h is the theoretical power output of renewable unit h.

[0034] 7) Power grid transmission security constraints: ; where GSF l-i , GSF l-h and GSF l-k are the power transfer factors of thermal unit i, renewable unit h and load k to transmission equipment l, respectively, F l,min and F l,max are the power flow and its upper and lower limits of line i-j, respectively.

[0035] Step S3: Based on the basic data of the entire market, and taking into account the start-up schedule and real-time market source-grid-load data, construct an economic dispatch model that takes into account the power grid security constraints.

[0036] The objective function of the economic dispatch model considering power grid security constraints is: ; In the formula, NG represents the number of conventional units in the system; p i and p h C represents the active power of conventional unit i and renewable energy unit h, respectively; i (p i ) and C h (p h ) represents the operating costs of units i and h. The conventional unit cost model includes carbon emission costs.

[0037] In addition, the model also needs to satisfy other constraints: 1) System load balance constraints: ; In the formula, D k NLD represents the number of load buses.

[0038] 2) Maximum and minimum generator output constraints: ; In the formula, pimin and pimax are the upper and lower limits of the output power of generator set i, respectively.

[0039] 3) Gradient Ratio Constraints: ; In the formula, pi is the maximum value of the load that unit i can add or remove in each time period.

[0040] In addition, the objective function also needs to satisfy the unit spinning reserve capacity constraint, renewable energy generation constraint, and grid transmission security constraint, with the same constraints as the expression in SCUC.

[0041] Step S4: Calculate the nodal marginal electricity price and power flow based on the constructed unit combination model and economic dispatch model that consider grid security constraints.

[0042] After the market SCED calculations were completed, for units that cannot be priced, the upper and lower limits of their unit output in the SCED model were replaced with the following fixed output constraints: ; in, This refers to the winning bid output of unit i during time period t, based on the SCED calculation results from the market a few days ago.

[0043] After fixing the output of non-pricing units in the corresponding period, the SCED model in the day-ahead market is recalculated to obtain the Lagrange multipliers of each period, and the node price of node k in period t is: ; wherein, is the Lagrange multiplier of the system load balance constraint in period t; and are the Lagrange multipliers of the maximum forward and reverse power flow constraints of branch l, respectively, and when the branch power flow is out of limit, the Lagrange multiplier is the network power flow constraint relaxation penalty factor; and are the Lagrange multipliers of the maximum forward and reverse power flow constraints of section s, respectively, and when the section power flow is out of limit, the Lagrange multiplier is the network power flow constraint relaxation penalty factor; and is the generator output power transfer distribution factor of node k to section s. The annual marginal price result obtained according to the calculation process is shown in Figure 3 .

[0044] At the same time, the change factors of the annual time power of new energy can be considered to construct the wind power time power curve and the photovoltaic annual time power curve, and the results are shown in Figure 4 、 Figure 5 .

[0045] Step S5, the result information is output based on the node price calculation result, and the uncertainty factors existing in the electricity market are evaluated.

[0046] The typical simulation calculation and its power supply income evaluation is a "most likely" scenario, that is, according to the current electricity market and the current available macro and enterprise planning conditions, the basic example simulation data is constructed. The uncertainty analysis needs to include macro factors and micro factors.

[0047] Step S6, based on the results of the uncertainty analysis, the statistics of various technical performances and economic indicators of the system are analyzed. The annual 8760-hour spot market simulation is carried out, and finally the simulation price obtained by the analysis is compared with the actual price, and it can be found that the two basically coincide. Details are shown in Figure 6 、 Figure 7 、 Figure 8 .

[0048] On the basis of the constructed electricity market, combined with the relevant planning information of the state and the local, different electricity market simulation models are constructed, and the data of conventional power, new energy, and power load are adjusted accordingly to construct the basic simulation example. The results obtained by the final simulation can obtain the influence of the change of various factors on the change of the price.

[0049] The present application encompasses any alternatives, modifications, equivalent methods and solutions made to the essence and scope of the present application. In order to make the public have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can also be fully understood without the description of these details to those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits, etc. are not described in detail.

[0050] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can also be made, which should be considered as the protection scope of the present application.

Claims

1. A medium- to long-term price forecasting method based on long-term electricity market simulation, characterized in that, Includes the following steps: S1, Market Basic Data Acquisition: Acquire basic market data information, including power supply side data, load side data, and fuel side data; S2, Unit Combination Model Construction and Day-ahead Simulation: Based on the basic data information of the entire market, and taking into account various day-ahead forecasts and maintenance arrangements, a unit combination model considering grid security constraints is constructed, and day-ahead simulation is performed; S3, Economic Dispatch Model Construction: Based on the basic data information of the entire market, and taking into account the start-up schedule and real-time market source-grid-load data, an economic dispatch model considering grid security constraints is constructed. S4, Node Marginal Electricity Price and Power Flow Calculation: Node marginal electricity price and power flow calculation are performed based on the constructed unit combination model considering grid security constraints and the economic dispatch model considering grid security constraints. S5, Uncertainty Factor Assessment: Based on the calculation results of the nodal marginal electricity price, assess the uncertainties existing in the electricity market; S6, Technical Performance and Economic Indicator Analysis: Based on the results of the uncertainty factor assessment, statistical analysis is performed on various technical performance and economic indicators of the system.

2. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 1, characterized in that, The acquisition of market-based data in S1 includes: S11, Power supply side data acquisition: By constructing a photovoltaic power generation output simulation model, based on solar irradiance and installed capacity, a photovoltaic power generation curve of 8760 hours in historical years is generated, and the curve is adjusted proportionally according to the predicted installed capacity of the target planning year to obtain photovoltaic power generation output data for the planning year. S12, Load-side data acquisition: Based on historical load data, the daily peak load for the whole year is calculated using interpolation, and the actual changes are simulated by random noise to generate 8760 hours of historical load data. The load data is then adjusted according to the peak load and total electricity demand forecast for the planned year to obtain a load curve that meets the demand for the planned year.

3. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 2, characterized in that, The power-side data acquisition in S11 includes: S111, Constructing the Daily Photovoltaic Power Generation Characteristic Curve: This study investigates the solar irradiance and sunshine hours of a local photovoltaic power station. Assuming the power generation capacity of the photovoltaic power station is m (ten thousand kilowatts), the daily photovoltaic power generation characteristic curve for this region is constructed based on the solar irradiance and sunshine hours, expressed as: S i (i=1,2,...24); S112, Constructing the monthly maximum photovoltaic power generation curve: Based on the changing characteristics of solar irradiance and sunshine duration in the four seasons (longer sunshine duration in summer and shorter sunshine duration in winter), a monthly maximum photovoltaic power generation curve is constructed, represented as: S j (j=1,2,...12); S113, Constructing the historical annual time-series output curve of solar power generation: Combining the daily photovoltaic power generation characteristic curve and the monthly maximum photovoltaic power generation output curve, constructing the 8760-hour power generation output curve of the photovoltaic power station in historical years. S114, Target Planning Year Photovoltaic Power Generation Output Curve Prediction: The predicted installed capacity of the photovoltaic power station in the target planning year is set as n (ten thousand kilowatts). Assuming no change in the region's solar irradiance characteristics until the target planning year, the historical 8760-hour power generation output curve is adjusted according to the ratio of the power station's installed capacity to its total power generation capacity in the target planning year. This yields the 8760-hour power generation output curve for the photovoltaic power station in the target planning year, expressed as: ； Among them, P t,plan.solar (10,000 kilowatts) represents the photovoltaic power generation capacity at time t in the target year, P t,hist.solar (10,000 kW) represents the photovoltaic power generation at time t in historical year, m (10,000 kW) represents the installed capacity of the solar power plant, and n (10,000 kW) represents the predicted installed capacity of the solar power plant in the target planning year.

4. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 3, characterized in that, The load-side data acquisition in S12 includes: S121, Construct the annual load curve for the planning year: Obtain historical monthly peak load data M for a certain region or province. j (j=1,2...12) and daily (winter and summer) 24-hour load data W for a certain region or province. n and S n (n=1,2,3...24); S122, Calculate the peak daily load data for the whole year: Calculate the peak daily load data Di (i=1,2...365) for the whole year using the interpolation method. S123, Adding noise to simulate the actual load curve: Noise is added to the annual daily load peak data Di using random numbers to obtain 365 daily load curves that are close to the actual load, forming 8760 load data for historical years; S124, calculate historical 8760-hour load data: Let S be the daily 24-hour load data for summer (May-September) and winter (October-April). n and W n Extract the maximum load demand within 24 hours of the day as the daily load peak H. i ,according to and The solution yields new 24-hour load data for winter and summer, thus providing 24-hour load data for 365 days, which is equivalent to obtaining the historical 8760-hour load data H for each year. k (k=1,2...8760); S125, Target Planning Year Load Curve Forecast: Historical Annual Load Peak Value Set as D max The peak load forecast for the planned year is set as follows: By adjusting proportionally, the planned annual load data for 8760 hours can be obtained. ; S126, Adjust the load curve to match the total electricity demand of the planned year: Given the total electricity consumption C of historical years and the predicted electricity demand C' of the planned year, keep the peak load of the planned year unchanged, adjust the 8760-hour load curve according to the deviation degree so that the total electricity demand of its annual continuous load curve meets the requirements of the predicted electricity demand of the target planned year, and calculate the 8760-hour load curve of the planned year.

5. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 4, characterized in that, The objective function of the unit combination model considering grid security constraints is expressed as: ； Among them, F[p i (t),I i [t] represents the total power generation cost of the system, p i (t) and I i (t) is the decision variable, p i (t) represents the output of unit i during time period t. i (t) represents the operating state of unit i during time period t. i (t)=1 indicates that the unit is in operation, I i (t)=0 indicates that the unit is in a stopped state, C i [p i [(t)] represents the power generation and operating cost of unit i during time period t, S i (t) represents the startup cost of unit i in time period t, M represents the number of units, and T represents the total number of time periods covered by the unit combination problem; The constraints of the unit combination model considering grid security constraints include: System power balance constraints: ; Among them, P d (t) represents the total load of the system during time period t; Maximum and minimum output constraints for generators: ; Among them, P imin P imax These are the minimum and maximum output of unit i, respectively; Unit spinning reserve capacity constraints: ; Among them, S D (t) represents the total reserve capacity requirement of the system during time period t; Slope rate constraint: ; Among them, DR i DR i These represent the allowable upward and downward output adjustments for unit i in each time period; Minimum operating and shutdown duration constraints for the unit: ； ； in, , For unit i, the shortest operating and downtime. , The continuous start-up and shutdown time of unit i before time period t; Renewable energy generation constraints: ; Among them, P h Theoretically, this will provide power to the renewable energy unit h. Power grid transmission safety constraints: ; Among them, GSF l-i GSF l-h GSF l-k The generation transfer factors of thermal power unit i, new energy unit h, and load k to transmission equipment l are respectively, F l,min and F l,max These represent the power flow and upper and lower limits of line ij, respectively.

6. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 5, characterized in that, The objective function of the economic dispatch model considering power grid security constraints is expressed as follows: Where NG represents the number of conventional units in the system, p i and p h The active power of conventional unit i and renewable energy unit h are respectively, C i (p i ) and C h (p h ) represents the operating costs of units i and h; The constraints of the economic dispatch model considering power grid security constraints include: System load balancing constraints: ; Among them, D k NLD represents the number of load buses; Maximum and minimum output constraints for generators: ; Where pimin and pimax are the upper and lower limits of the output power of generator set i, respectively; Slope rate constraint: ; Where pi is the maximum value that unit i can add or remove load in each time period.

7. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 6, characterized in that, The nodal marginal electricity price and power flow calculation in S4 include: S41, Handling the output power of non-priceable generating units: For non-priceable generating units, in the economic dispatch model considering grid security constraints, the upper and lower limits of their unit output are replaced with fixed output constraints, expressed as: ； in, The winning bid output of unit i in time period t, based on the SCED calculation results of the market in the previous day; S42, Calculate the nodal marginal price: After fixing the output of non-pricing units, recalculate the economic dispatch model in the day-ahead market considering grid security constraints, obtain the Lagrange multipliers for each time period, and calculate the nodal price of node k in time period t, expressed as: ； in, For the system load balance constraint in time period t, the Lagrange multiplier is... and These are the Lagrange multipliers for the maximum forward and maximum reverse power flow constraints of branch l, respectively. Furthermore, when the branch power flow exceeds the limit, these Lagrange multipliers become the network power flow constraint relaxation penalty factor. and These are the Lagrange multipliers for the maximum forward and maximum reverse power flow constraints of section s, respectively. Furthermore, when the power flow of a section exceeds its limits, these Lagrange multipliers become the network power flow constraint relaxation penalty factor. Let be the generator output power transfer distribution factor at node k to section s.

8. The medium- to long-term price forecasting method based on long-term electricity market simulation according to claim 7, characterized in that, The uncertainties in S5 include both macroscopic and microscopic uncertainties.

Images