A method for optimizing the configuration of heterogeneous power sources in new energy bases to adapt to multiple operating conditions
By constructing a set of multiple operating conditions and improving the multi-objective particle swarm optimization algorithm, the installed capacity of wind, solar, thermal and energy storage and the converter ratio of the new energy base were optimized, which solved the stability and absorption capacity problems of the new energy base under weak grid conditions and achieved efficient power supply configuration and economic operation.
Patent Information
- Application Number
- CN202611627376.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing power configuration methods for new energy bases suffer from insufficient consideration of stability constraints, limited scenario coverage, lack of optimization of grid connection/network mode combinations, prominent contradiction between solution efficiency and optimization accuracy, and imperfect multi-objective coordination mechanisms. These issues result in insufficient stability and limited absorption capacity of new energy bases under weak grid conditions.
A typical power generation time series scenario set covering different short-circuit ratio levels and multiple operating condition combinations is constructed. A heterogeneous power source optimization configuration model is established, and a matching index based on grid strength assessment is introduced. An improved multi-objective particle swarm optimization algorithm is adopted to perform rapid stability pre-screening and multi-objective optimization, optimize the installed capacity and converter matching of wind, solar, thermal and energy storage, and achieve coordination of system stability, new energy consumption capacity and external transmission economy.
It improves the robustness of new energy bases under various complex operating conditions, enhances the ability of new energy to be transmitted to other regions in weak grid scenarios, shortens the optimization solution time, reduces the overall cost, and improves the utilization rate of new energy and the voltage recovery speed.
Smart Images

Figure CN123121817A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of new energy power generation planning and power system optimization scheduling technology, specifically, to a method for optimizing the configuration of heterogeneous power sources in new energy bases that adapts to multiple operating conditions. Background Technology
[0002] With the deepening of the goal of "carbon peaking and carbon neutrality", the construction scale of large-scale wind and solar power bases in my country, focusing on desert, Gobi and arid areas, continues to expand. Large-scale new energy bases mainly take the form of wind, solar, hydro and storage, and wind, solar, thermal and storage multi-energy complementarity, and transmit and consume them through ultra-high voltage DC and AC channels.
[0003] In large-scale new energy bases, the proportion of new energy installed capacity is relatively high, resulting in strong random fluctuations in power output. Furthermore, the power is primarily transmitted over long distances to load centers in central and eastern China for consumption. This leads to a dual-high characteristic of the sending-end power grid: a high proportion of new energy and a high proportion of power electronic equipment. Consequently, the system voltage support strength decreases, and the short-circuit ratio continues to decline. Under weak or even extremely weak grid conditions, transient voltage instability caused by phase-locked loops in grid-connected converters becomes particularly prominent, affecting the safe and stable operation of the system.
[0004] Existing technical solutions for power capacity configuration in new energy bases have the following main shortcomings:
[0005] First, stability constraints are not adequately considered. Existing capacity configuration methods mostly focus on economic efficiency and absorption capacity assessment, and the quantitative analysis of system voltage support strength and short-circuit ratio constraints is not in-depth enough. The short-circuit ratio is usually used as a post-verification indicator rather than an optimization constraint, which may lead to insufficient stability of the configuration scheme under weak grid conditions.
[0006] Secondly, the scope of scenario coverage is limited. Existing methods mostly use typical daily or single-year time-series scenarios for production simulation, without fully considering the complex scenario coverage requirements of strong, weak, and extremely weak short-circuit ratios and multiple operating conditions. Traditional production simulation based on an annual time series of 8760 hours involves a large amount of computation and is difficult to effectively cover extreme operating conditions.
[0007] Third, there is a lack of optimization for the combination of grid-connected and grid-connected modes. Existing configuration methods usually treat the converter control mode of new energy power plants as a fixed parameter, and do not include the ratio of grid-connected and grid-connected converters as an optimization variable in the capacity configuration model. Under weak grid conditions, grid-connected converters are prone to transient voltage instability caused by phase-locked loops during transient periods, which urgently needs to be addressed through optimal matching.
[0008] Fourth, there is a significant contradiction between solution efficiency and optimization accuracy. If each candidate solution is simulated and verified in the full time domain during the capacity optimization process, the computational load increases exponentially with the number of scenarios, and the solution efficiency is difficult to meet engineering requirements.
[0009] Fifth, the multi-objective coordination mechanism is imperfect. There are coupled and contradictory relationships among multiple objectives such as investment and operating costs, dynamic voltage support capacity, and renewable energy utilization rate. It is practically difficult to deploy large-scale distributed synchronous condensers to ensure that the short-circuit ratio of multiple renewable energy power plants meets the requirements due to high costs. Summary of the Invention
[0010] To address the technical problems of existing power configuration methods for new energy bases, such as insufficient consideration of stability constraints, limited scenario coverage, lack of optimization of grid connection / network configuration combinations, significant contradiction between solution efficiency and optimization accuracy, and imperfect multi-objective coordination mechanisms, this invention provides a method for optimizing the configuration of heterogeneous power sources in new energy bases that is adaptable to multiple operating conditions. This method constructs a set of typical power generation output time-series scenarios covering different short-circuit ratio levels and multiple operating condition combinations. It establishes an optimization configuration model that includes the capacity ratio of heterogeneous power sources (wind, solar, thermal, and energy storage) and grid connection / network configuration combinations. Furthermore, it introduces a ratio index based on grid strength assessment for rapid stability pre-screening and employs an improved multi-objective particle swarm optimization algorithm for solution. This allows for the rapid solution of optimal ratio schemes for multiple types of power sources while considering system stability, new energy absorption capacity, and external transmission economics.
[0011] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0012] A method for optimizing the configuration of heterogeneous power sources in a new energy base to adapt to multiple operating conditions includes the following steps:
[0013] Step 1: Construct a set of typical scenarios with multiple operating conditions: Collect wind and solar power output data and grid operation boundary conditions from new energy bases, and construct a set of typical power generation output time series scenarios covering different short-circuit ratio levels and multiple operating condition combinations;
[0014] Step 2: Establish a heterogeneous power source optimization configuration model: Using the installed capacity of wind power, photovoltaic power, thermal power, and energy storage, the energy storage power ratio, and the matching coefficient of grid-connected and grid-connected converters of each new energy power station as decision variables, and taking comprehensive investment and operating costs, dynamic voltage support capacity, and new energy utilization rate as multiple objectives, and setting power balance constraints, multi-station short-circuit ratio constraints, conventional unit operation constraints, energy storage operation constraints, DC transmission channel constraints, and grid-connected / grid-connected matching constraints, an optimization configuration model is established.
[0015] Step 3: Rapid stability verification: Introduce a ratio index based on power grid strength assessment to perform rapid stability pre-screening of candidate configuration schemes, and only perform time-domain simulation verification on candidate configuration schemes that are in the critical range of the pre-screening.
[0016] Step 4: Optimization and Solution: An improved multi-objective particle swarm optimization algorithm is adopted, with power capacity configuration as the outer layer and operation verification as the inner layer, to solve the optimization configuration model and obtain the optimal power configuration scheme. This enables the rapid solution of the best ratio scheme of multiple types of power sources while taking into account system stability, new energy absorption capacity and external transmission economy.
[0017] Specifically, step one includes the following process:
[0018] Collect meteorological data on wind speed, solar irradiance, and temperature, as well as historical power output data from wind and solar power plants, and collect grid operation boundary conditions;
[0019] Based on the short-circuit ratio at the grid connection point, the power grid strength is divided into three levels: strong grid, weak grid, and extremely weak grid. Combined with load level, transmission channel operation mode, and conventional unit start-up mode, a three-dimensional operating condition combination matrix of short-circuit ratio level × load level × transmission mode is constructed.
[0020] Furthermore, in step one, considering the spatiotemporal correlation between wind power output and photovoltaic power output, a joint probability distribution model of wind power output and photovoltaic power output is constructed using a Copula function, and a large-scale initial scene set is generated using Latin hypercube sampling; the initial scene set is simplified using a scene reduction method based on Kantorovich distance to form the typical scene set.
[0021] Furthermore, the scenario reduction adopts the synchronous back-substitution elimination method to gradually eliminate the scenario pairs with the smallest probability distance and merge their probabilities into the retained scenarios; the typical scenario set formed has a scale of 20 to 50 scenarios, each scenario contains the wind and solar power output time series curves on a 24-hour or longer time scale, and is labeled with the corresponding short-circuit ratio level and operating condition label.
[0022] Specifically, in step two, the energy storage power-to-energy ratio is defined as the ratio of energy storage power capacity to energy storage capacity; and the energy storage duration is defined as the ratio of energy storage capacity to energy storage power capacity.
[0023] Specifically, in step two, the matching ratio coefficient between grid-type and grid-connected converters is defined as the proportion of the capacity of grid-connected converters to the total converter capacity of the station, with a value range of 0 to 1.
[0024] Specifically, in step two, the multi-objective includes three sub-objectives: minimizing the overall investment and operating cost, maximizing the dynamic voltage support capacity, and maximizing the utilization rate of new energy sources. The three sub-objectives are normalized and then weighted and summed using the weight coefficient method to transform them into a comprehensive objective function. Among them, the dynamic voltage support capacity is represented by the sum of the dynamic voltage support capacity indicators of the system under each scenario according to the scenario probability weight, and the new energy utilization rate is represented by the ratio of the sum of the actual wind power and photovoltaic power consumed in each time period under each scenario to the sum of the theoretical available power.
[0025] Specifically, in step two, the short-circuit ratio constraint for multiple power plants is determined by the short-circuit capacity of the power plant, the active power output of the new energy power plant, and the mutual impedance and self-impedance of the system impedance matrix, and is not less than the minimum short-circuit ratio threshold. After configuring grid-type converters or distributed synchronous condensers, the short-circuit ratio of the multiple power plants is increased by increasing the system short-circuit capacity.
[0026] Specifically, in step three, the matching ratio index is determined by the sum of the basic short-circuit ratio and the equivalent short-circuit ratio increment of the grid-type converter. The basic short-circuit ratio is the ratio of the system's basic short-circuit capacity excluding the contribution of the grid-type converter to the total apparent power of new energy sources. The equivalent short-circuit ratio increment is determined by the grid-type converter matching ratio coefficient and the equivalent enhancement coefficient of the short-circuit ratio per unit capacity of the grid-type converter. The matching ratio index increases with the increase of the grid-type converter matching ratio coefficient. When the matching ratio index of a candidate configuration scheme is greater than the preset threshold of the corresponding short-circuit ratio level, the candidate configuration scheme is determined to meet the stability requirements. When the matching ratio index is within a preset critical range, time-domain simulation verification is triggered. When the matching ratio index is lower than the preset lower limit of the critical range, the candidate configuration scheme is determined to not meet the stability requirements and is directly eliminated.
[0027] Furthermore, in step four, the improved multi-objective particle swarm optimization algorithm introduces an adaptive inertial weight mechanism, an elite historical experience reuse strategy, a dual-mode oscillation disturbance mechanism, and a constraint domination principle on the basis of the standard multi-objective particle swarm optimization algorithm.
[0028] The outer layer uses the ratio of wind, solar, thermal and energy storage capacity and the ratio of root / structure network as optimization variables to perform a global search. The inner layer uses a mixed integer linear programming model to perform operation verification in various typical scenarios, and the operation verification results of the inner layer are fed back to the outer layer as the basis for fitness evaluation.
[0029] After obtaining the Pareto optimal solution set, a multi-criteria decision-making method is used to select the best compromise solution.
[0030] Compared with the prior art, the present invention has the following beneficial effects:
[0031] (1) By constructing a typical power generation output time series scenario set covering strong, weak and extremely weak short-circuit ratio levels and multiple operating conditions, this invention enables the configuration scheme to maintain robustness under various complex operating conditions, thereby overcoming the limitations of existing methods that only consider typical daily or single-year scenarios and improving the adaptability of the configuration scheme to the actual operating boundary conditions of new energy bases.
[0032] (2) This invention directly embeds the short-circuit ratio constraint of multiple stations into the optimization configuration model and uses the equivalent improvement mechanism of short-circuit capacity of network-type equipment to achieve the coordinated optimization of stability constraints and capacity configuration. In typical cases, the new energy transmission capacity in weak grid scenarios can be improved by about 10% to 15%.
[0033] (3) This invention introduces a ratio index based on power grid strength assessment for rapid stability pre-screening, avoiding scenario-by-scenario full-time domain simulation, thereby shortening the optimization solution time by more than 60% while maintaining the same configuration accuracy as the full simulation method.
[0034] (4) The present invention coordinates and optimizes investment and operation costs, dynamic voltage support capability and new energy utilization rate through a multi-objective optimization framework, thereby increasing the new energy utilization rate by about 7.3 percentage points, reducing the overall cost by about 10.3%, and shortening the transient voltage recovery time in extremely weak power grid scenarios by about 42% in the embodiments. Attached Figure Description
[0035] Figure 1 This is a schematic diagram of the overall technical process of an embodiment of the present invention.
[0036] Figure 2 This is a schematic diagram illustrating the process of constructing a set of typical operating conditions in an embodiment of the present invention.
[0037] Figure 3 This is a schematic diagram of the heterogeneous power supply optimization configuration model framework in an embodiment of the present invention.
[0038] Figure 4 This is a schematic diagram of a rapid stability verification process based on proportion indicators in an embodiment of the present invention.
[0039] Figure 5 This is a schematic diagram of the improved multi-objective particle swarm optimization algorithm in an embodiment of the present invention.
[0040] Figure 6 This is a schematic diagram of a two-layer optimization solution architecture in an embodiment of the present invention.
[0041] Figure 7 This is a schematic diagram of the Pareto front and preferred results in an embodiment of the present invention.
[0042] Figure 8This is a comparison diagram of the transient voltage response of the system under different root / network ratio schemes in an embodiment of the present invention. Detailed Implementation
[0043] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.
[0044] Example
[0045] like Figures 1 to 8 As shown, this embodiment provides a method for optimizing the configuration of heterogeneous power sources in a new energy base to adapt to multiple operating conditions, which mainly includes the following four steps:
[0046] Step 1: Construct a set of typical scenarios with multiple operating conditions: Collect wind and solar power output data and grid operation boundary conditions from new energy bases, and construct a set of typical power generation output time series scenarios covering different short-circuit ratio levels and multiple operating condition combinations.
[0047] Step 2: Construct a heterogeneous power source optimization configuration model: Using the installed capacity of wind power, photovoltaic power, thermal power, and energy storage, the energy storage power ratio, and the matching coefficient of grid-connected and grid-connected converters of each new energy power station as decision variables, and taking comprehensive investment and operating costs, dynamic voltage support capacity, and new energy utilization rate as multiple objectives, and setting power balance constraints, multi-station short-circuit ratio constraints, conventional unit operation constraints, energy storage operation constraints, DC transmission channel constraints, and grid-connected / grid-connected matching constraints, an optimization configuration model is established.
[0048] Step 3, rapid stability verification based on power grid strength assessment: Introduce a matching index based on power grid strength assessment to perform rapid stability pre-screening of candidate configuration schemes, and perform time-domain simulation verification only on candidate schemes that fail the pre-screening or are in the critical range.
[0049] Step 4, Improve the intelligent optimization algorithm to solve: Use an improved multi-objective particle swarm optimization algorithm, with power capacity configuration as the outer layer and operation verification as the inner layer, to solve the optimization configuration model and obtain the optimal power configuration scheme.
[0050] The details are as follows:
[0051] In step one, meteorological data such as wind speed, solar irradiance, and temperature in the area where the new energy base is located, as well as historical output data of wind power and photovoltaic power stations, are collected. The time resolution is no less than 1 hour, and the historical data period is no less than 3 years. At the same time, the power grid operation boundary conditions are collected, including the topology and line parameters of the sending-end power grid, the installed capacity and operating characteristic parameters of conventional units, the configuration parameters of energy storage systems, the transmission capacity constraints and regulation characteristics of DC transmission channels, and the load characteristic data of the receiving-end power grid. Outlier removal, missing value imputation, and normalization are performed on the collected data to form a standardized input dataset.
[0052] Then, based on the short-circuit ratio at the grid connection point, the grid strength is divided into three levels: strong grid, weak grid, and extremely weak grid. Combining this with multi-dimensional operating conditions such as system load level, transmission channel operation mode, and conventional unit start-up mode, a three-dimensional operating condition combination matrix of "short-circuit ratio level × load level × transmission mode" is constructed. Specifically, the set of short-circuit ratio levels is denoted as... These correspond to SCR≥3, 2≤SCR<3, and SCR<2, respectively; let the load level set be denoted as . These correspond to peak, average, and off-peak periods, respectively; the set of delivery methods is denoted as... These correspond to full delivery, minimum order, and service suspension, respectively. The Cartesian product of the above three sets... This constitutes a three-dimensional working condition combination matrix, totaling... Each operating condition unit uses a triplet. The logo, among which , , .
[0053] Based on this, the historical fluctuation characteristics of new energy sources are considered to correct the time-series sampling probability interval of the prediction error. Prediction error probability distribution models are established for wind power output and photovoltaic power output respectively, and normal or Beta distributions are used to fit the prediction errors for each time period. Specifically, the set of historical sample time periods on which the scene generation is based is denoted as... , Indexed by historical time period; wind power and photovoltaic power in historical time period t h The predicted outputs are respectively , The actual outputs were respectively , Define the prediction error of wind power. Prediction error of photovoltaics They are respectively:
[0054]
[0055] To eliminate the influence of the dimensions of installed capacity, normalized prediction error is adopted:
[0056]
[0057] In the above formula, C W and C PV These refer to wind power installed capacity and photovoltaic installed capacity, respectively.
[0058] Normalized prediction error for wind power The marginal distribution is fitted using a normal distribution:
[0059]
[0060] In the above formula, Let be the cumulative distribution function of the normalized prediction error of wind power, representing the probability that the normalized prediction error of wind power is ≤ x, and the independent variable x is the value of the normalized prediction error of wind power. and These represent the mean and standard deviation of the normalized prediction error for wind power, respectively. It is the standard normal distribution function.
[0061] Normalized prediction error of photovoltaic Since its values are usually limited to the interval [-1, 1] and its distribution may be skewed, a Beta distribution is used to fit its marginal distribution:
[0062]
[0063] In the above formula, Let be the cumulative distribution function of the photovoltaic normalized prediction error, representing the probability that the photovoltaic normalized prediction error is ≤ y, and the independent variable y is the value of the photovoltaic normalized prediction error. and Let be the shape parameter of the Beta distribution. This is a regularized incomplete Beta function; through variable transformation. Will Mapped to the domain [0,1] of the Beta distribution. When the photovoltaic prediction error distribution is approximately symmetrical, it can also degenerate into a normal distribution fit. The above marginal distribution parameters... , , , All scenarios were determined using the maximum likelihood estimation method based on historical samples. A large-scale initial scene set of no fewer than 2000 scenarios was generated by stratified random sampling from the probability distribution function using the Latin hypercube sampling method.
[0064] Considering the spatiotemporal correlation between wind power output and photovoltaic power output, a joint probability distribution model is constructed using a copula function; specifically, the joint distribution function of the normalized prediction errors of wind power and photovoltaic power is set as follows: According to Sklar's theorem, there exists a connection to the Copula function. Make:
[0065]
[0066] in These are the parameters for connecting the Copula function. In this embodiment, Gaussian Copula or t-Copula can be selected as candidate functions, based on historical power output samples. First, estimate the marginal distribution parameters, then perform a probability integral transformation on the samples to obtain... ,in , These are the quantiles of the normalized prediction error samples for wind power and photovoltaic power, respectively, after probability integral transformation, and they follow a uniform distribution in [0,1]. The Copula parameters are estimated using the maximum likelihood estimation method. Calculate the AIC and BIC indices for Gaussian Copula and t-Copula respectively:
[0067]
[0068] In the above formula, L is the likelihood function value, k is the number of model parameters, and Gaussian Copula takes... t-Copula ; The number of historical samples is used. The type with the smaller AIC and BIC is selected as the optimal Copula type.
[0069] Subsequently, Latin hypercube sampling is performed on the selected Copula function to generate a correlated uniform random number sample. Then, by inverse transformation of the marginal distributions of each variable, a joint time-series sample of wind power and photovoltaic output is obtained:
[0070]
[0071] This yields the wind power and solar power output samples at time t under scenario s:
[0072]
[0073] In the above formula, , The prediction output is for the time period t in which the scene to be generated is to be generated; the samples also need to be calculated according to... and The range is truncated and corrected to ensure that the output is non-negative and does not exceed the installed capacity.
[0074] Finally, a scenario reduction method based on Kantorovich distance was used to simplify the initial scenario set. The Kantorovich probability distance between each scenario pair was calculated, and the scenario pair with the smallest probability distance was gradually eliminated using a synchronous back-substitution elimination method. Their probabilities were then merged into the retained scenarios. The reduced scenarios were mapped and matched with the three-dimensional operating condition combination matrix to form a typical power generation time series scenario set covering strong, weak, and extremely weak short-circuit ratios and multiple operating conditions. The size of the typical scenario set was controlled at 20-50 scenarios, and each scenario contained wind and solar power output time series curves on a 24-hour or longer time scale, with corresponding short-circuit ratio levels and operating condition labels.
[0075] In step two, a heterogeneous power source optimization configuration model is constructed. The decision variables of the optimization configuration model include: wind power installed capacity C. W Photovoltaic installed capacity C PV Thermal power installed capacity C TH Energy storage installed capacity C ES And the energy ratio of energy storage power λ ES And the matching ratio coefficient α of each new energy power station and grid-type and grid-type converter. GFM Among them, C ES Defined as energy storage capacity (MWh), P ES Defined as energy storage capacity (MW), energy storage power ratio λ ES =P ES / C ES Dimensionless -1 Energy storage duration T ES =C ES / P ES The dimensionless value is h, which represents the duration of the energy storage system charging and discharging at rated power. The proportioning coefficient α GFM Defined as the proportion of grid-type converter capacity to the total converter capacity of the power station, with a value range of 0 ≤ α. GFM ≤1.
[0076] Regarding the objective function, a multi-objective optimization function containing three sub-objectives is constructed:
[0077] First, minimize the overall investment and operating cost F1:
[0078]
[0079] In the above formula, C inv To calculate the annualized investment cost, considering the unit capacity investment cost and discount rate for various types of power sources, C... om For annual operation and maintenance costs, C fuel For the cost of fuel for thermal power, C pen The cost of curtailing wind and solar power. The economic penalty for curtailing wind and solar power is incorporated into the total life-cycle cost of electricity loss, and carbon trading costs can be dynamically adjusted according to carbon price levels.
[0080] Second, maximize the dynamic voltage support capability F2:
[0081]
[0082] In the above formula, VSF s This is a dynamic voltage support capability indicator for the system under scenario s, comprehensively reflecting the transient voltage recovery speed and steady-state voltage deviation at each grid connection point; w s The probability weights for scenario s are given. The voltage support capability index is calculated based on a combination of the voltage source characteristics of the grid-connected converter and the current source characteristics of the grid-connected converter.
[0083] Third, maximize the utilization rate of new energy sources (F3):
[0084]
[0085] In the above formula, and These represent the actual wind power and photovoltaic power absorbed at time t under scenario s, respectively. and This corresponds to the theoretically available power; To prevent small positive numbers with a denominator of zero, when the theoretical available power of a certain scenario is zero throughout the entire period, that scenario will not participate in the F3 calculation, or the F3 for that scenario will be set to 1.
[0086] The three sub-objectives are transformed into a comprehensive objective function using the weighting coefficient method:
[0087]
[0088] In the above formula, , and The benchmark values for each sub-target (i.e., comprehensive investment and operating cost, dynamic voltage support capacity, and new energy utilization rate); , and For the corresponding weight coefficients, satisfying The weighting coefficients can be determined using the analytic hierarchy process (AHP) or the entropy weighting method. Each benchmark value can be taken as the corresponding index value of the initial planning scheme, or as the average value of each sub-objective sample within the optimization search space. When the benchmark value of a sub-objective is zero, the maximum or minimum value of that sub-objective in the feasible solution set is used instead to ensure that the normalization term is defined.
[0089] Regarding constraints, the following constraints are set:
[0090] (1) Power balance constraint. Let the active power output of wind power, photovoltaic power, thermal power and energy storage discharge at time t under scenario s be respectively , , , The active power of local load, DC transmission, and energy storage charging are respectively , , Then the power balance constraint is expressed as:
[0091]
[0092] In the above formula, the total output of thermal power plants satisfies Where g is the thermal power unit number. For thermal power units; the total DC power transmission meets the requirements. Where d is the DC transmission channel number, This is a collection of DC power transmission channels; For the set of scheduling periods, This is a collection of typical scenarios.
[0093] (2) Short-circuit ratio constraint for multiple new energy power plants. Assume there are n new energy power plants within the new energy base, each connected to the grid via its own connection point. The short-circuit capacity of the connection point of new energy power plant i is... The active power output of station i is P i The active power output of all stations except station i is P. j The self-impedance of the system impedance matrix is Z. ii mutual impedance Z ij Then the multi-station short-circuit ratio constraint of new energy power station i is expressed as:
[0094]
[0095] In the above formula, MRSCR i For station i, the multi-station short-circuit ratio. P is the short-circuit capacity of the grid connection point of station i. i For the active power contribution of station i, P j Z represents the active power output of all other power stations (excluding power station i) j. The summation term represents the coupling effect of other power stations on the short-circuit ratio of power station i at the grid connection point through mutual impedance. ii Z represents the self-impedance of the field station i at the grid connection point in the system impedance matrix. ij For the system impedance matrix, the mutual impedance between field station i and field station j at the grid connection point is MRSCR. min This is the minimum short-circuit ratio threshold for multiple stations.
[0096] When a grid-connected converter is configured, the grid-connected equipment increases the system's short-circuit capacity, thereby improving the short-circuit ratio at the renewable energy grid connection point. When configuring distributed synchronous condensers and grid-connected energy storage, the short-circuit ratio constraint can be transformed into a linear expression using the corresponding grid connection point short-circuit capacity modeling method, facilitating optimization. Furthermore, according to industry standards regarding the impedance ratio X / R criterion: when X / R < 10, the short-circuit ratio of multiple power plants at the generator unit connection point should be used; when X / R ≥ 10, the short-circuit ratio of multiple power plants at the power plant grid connection point should be used. This invention simultaneously calculates X / R at each grid connection point during constraint construction and selects the corresponding MRSCR calculation method accordingly to enhance engineering applicability.
[0097] (3) Conventional unit operation constraints, including upper and lower limits of thermal power unit output, ramp rate constraints and minimum start-up and shutdown time constraints.
[0098] The upper and lower limits of thermal power unit output are constrained as follows:
[0099]
[0100] In the above formula, Let g be the active power output of unit g at time t in scenario s. and These are the lower and upper limits of the output of unit g, respectively. This is a power-on status variable, where 1 indicates power-on and 0 indicates power-off. This refers to a collection of thermal power units.
[0101] The ramp rate constraint for thermal power units is:
[0102]
[0103]
[0104] In the above formula, and These represent the upper and lower ramp rates of unit g, respectively. The interval between adjacent time periods is t; when t=1, the unit output at the initial moment of scenario s is taken.
[0105] The minimum start-up and shutdown time constraint for thermal power units is defined as follows: Define the start-up operation variable of unit g at time t under scenario s. and shutdown operation variables ,satisfy:
[0106]
[0107]
[0108]
[0109]
[0110] In the above formula, Take 1 when the unit switches from shutdown to startup. Take 1 when the unit switches from startup to shutdown. and These represent the minimum continuous start-up time and minimum continuous shutdown time of unit g, respectively. The minimum start-up and shutdown time constraints for thermal power units can be flexibly selected based on the unit type and scheduling cycle during actual solution.
[0111] To ensure the grid strength of large-scale renewable energy grid-connected systems, conventional generating units are required to provide voltage support. However, excessively large operating capacity of conventional generating units will inevitably encroach on the space for renewable energy consumption. Therefore, it is necessary to optimize the minimum operating mode of generating units with the goal of maximizing renewable energy consumption, while meeting the maximum grid connection strength requirements of renewable energy equipment.
[0112] (4) Energy storage operation constraints, including energy storage charging and discharging power constraints, state of charge constraints and daily energy balance constraints.
[0113] In the energy storage charging and discharging power constraint, a binary variable is introduced. and Let represent the energy storage discharge and charging states, respectively, where 1 indicates that the energy is in that state and 0 indicates that the energy is not in that state. Then we have:
[0114]
[0115]
[0116]
[0117] In the above formula, This is the maximum rated discharge power of the energy storage. This is the maximum rated charging power for energy storage.
[0118] The state of charge constraint is:
[0119]
[0120]
[0121] In the above formula, For the energy storage state of charge at time t under scenario s, satisfying ; and These are the lower and upper limits of the energy storage state of charge, respectively; For storing energy at time t under scenario s, C ES Energy storage capacity; and These refer to energy storage charging efficiency and energy storage discharging efficiency, respectively. The interval between adjacent time periods; when t=1, Take the initial energy storage of scene s .
[0122] The daily energy balance constraint is:
[0123]
[0124] In the above formula, and These represent the stored energy at the initial and final moments of scenario s, respectively.
[0125] (5) DC transmission channel constraints, including the maximum transmission capacity constraint, daily adjustment frequency constraint and maximum power ramping capability constraint.
[0126] The maximum transmission capacity constraint for the external transmission channel is:
[0127]
[0128] In the above formula, Let be the transmission power of channel d at time t in scenario s. Let be the maximum transmission capacity of channel d. If the channel has bidirectional transmission capability, the constraint changes to: .
[0129] The maximum power ramping capability constraint is:
[0130]
[0131] In the above formula, This represents the maximum climbing capacity of channel d. Let be the transmission power of channel d at time t-1 in scenario s. When t=1, Let the DC transmission power be taken at the initial moment of scenario s. This constraint can be equivalently decomposed into two inequalities: upward ramp and downward ramp.
[0132]
[0133]
[0134] Define the adjustment action identifier variable of channel d at time t in scene s. The value is set to 1 when the power change in adjacent time periods exceeds the dead zone threshold, and 0 otherwise. A large M constant is introduced. Then the adjustment action identifier constraint is:
[0135]
[0136] The daily adjustment frequency constraint is:
[0137]
[0138] In the above formula, the large M constant can be taken as... , The dead zone threshold for channel d is adjusted. This is the daily adjustment limit for channel d. It is achieved by limiting the number of adjustments within a single scheduling day. The sum does not exceed This achieves a constraint on the number of daily adjustments.
[0139] (6) The meshing ratio constraint satisfies:
[0140]
[0141]
[0142] In the above formula, Minimum grid-type converter capacity required to meet transient voltage stability requirements; Let be the total capacity of the converters in power station i. The conversion relationship between this capacity and the installed capacity of new energy sources is as follows:
[0143]
[0144] In the above formula, The converter capacity configuration ratio is dimensionless and is typically taken as 1.0 to 1.2. The apparent power of new energy at station i; The installed capacity of new energy at station i; The equivalent power factor for new energy power plants is typically taken as 0.95 to 1.0. When taking... When, it can be simplified to: However, the conversion should be made according to the actual power factor during the project implementation to avoid mixing the units of MVA and MW.
[0145] The optimal matching strategy of grid-connected converters based on transient voltage stability constraints can ensure the stability of the PLL of the grid-connected converter under transient operating conditions. Combining the advantages of grid voltage support in the early stage of transient and grid voltage support in the stable period of transient, it enables new energy power plants to provide proactive, rapid and effective support for transient voltage.
[0146] In step three, a rapid stability check based on grid strength assessment is performed. To overcome the problem of excessive computational load in scenario-by-scenario full-time-domain simulation, a grid strength ratio index (GSRI) is introduced to quickly determine the stability and feasibility of candidate configuration schemes.
[0147] The GSRI metric is defined as follows:
[0148]
[0149] In the above formula, Based on the short-circuit ratio, S ac The system's basic short-circuit capacity includes only contributions from conventional generating units and synchronous condensers, excluding contributions from grid-connected converters. RE The total apparent power of new energy sources, and
[0150]
[0151] In the formula, This refers to the total active power of new energy sources.
[0152] This is the equivalent improvement factor of the short-circuit ratio per unit capacity for grid-type converters. ,in This is the overcurrent factor for a grid-type converter, typically ranging from 1.5 to 2.0. This is the impedance correction factor at the connection point, which can be determined based on the impedance characteristics of the connection point, the X / R ratio of the line, and the control strategy of the grid-type equipment. The typical value range is 0.8 to 1.0.
[0153] The physical basis for the above GSRI index is as follows: the contribution of grid-connected converters to short-circuit capacity mainly depends on their own overcurrent capacity and rated capacity, and is basically independent of the strength of the underlying power grid. Therefore, their contribution is manifested as a fixed increment to the base short-circuit ratio; when α GFM When = 0, GSRI degenerates into SCR. base .
[0154] When the GSRI index of the candidate configuration scheme is greater than the preset threshold GSRI of the corresponding short-circuit ratio level th When the GSRI index is within a preset critical range, detailed time-domain simulation verification is triggered. When the GSRI index is below the lower limit of the preset critical range, the scheme is determined to not meet the stability requirements at the corresponding short-circuit ratio level and is directly eliminated, not proceeding to the next step. The critical range is defined as follows: ,in The preset critical interval width can be taken as... , The preset interval coefficient can be set to 0.05 to 0.1. This mechanism can reduce the number of candidate solutions requiring full-time-domain simulation by more than 60%. Furthermore, the accuracy and recall of GSRI pre-screening can be statistically analyzed using historical simulation samples. When the pre-screening accuracy falls below 90% or the recall falls below 95%, GSRI is automatically tightened. th Or increase η GFM The calibration accuracy is improved to enhance the reliability of pre-screening.
[0155] GSRI threshold GSRI th Dynamic adjustment based on short-circuit ratio level and operating conditions: For each short-circuit ratio level, the minimum α required to meet transient voltage stability requirements is first determined through time-domain simulation. GFM Then, substitute the values into the GSRI index formula to calculate the corresponding GSRI value, which serves as the threshold for that level. Specifically, for a strong power grid (SCR≥3), the base short-circuit ratio is high, and the required GSRI is... th The incremental increase is relatively small, allowing for a higher penetration rate of new energy sources; for weak grids (2≤SCR<3), the required GSRI th A moderate increase in increments requires a certain grid configuration ratio; for extremely weak power grids (SCR < 2), the base short-circuit ratio is low, and the required GSRI is high. thThe incremental requirements are the largest, necessitating the configuration of a sufficient number of grid-connected converters and voltage support equipment. This embodiment illustrates the GSRI threshold calibration using the following short-circuit ratio levels: in a strong power grid Weak power grid extremely weak power grid In this scenario, take the typical value. , Typical values are obtained. The minimum grid configuration ratio α corresponding to this strong power grid scenario was determined through time-domain simulation of calibrated samples. GFM The minimum grid configuration ratio α corresponding to this weak grid scenario is approximately 0.06. GFM The minimum grid configuration ratio α corresponding to this extremely weak power grid scenario is approximately 0.19. GFM Approximately 0.31; calculated according to the GSRI index formula: the GSRI for this strong power grid scenario th The GSRI for this weak grid scenario is 3.6. th The GSRI for this extremely weak power grid scenario is 2.6. th The value is 2.0. For other operating conditions, the GSRI threshold corresponding to the short-circuit ratio level is calculated and determined based on the actual parameters and the time-domain simulation verification results of the calibration samples.
[0156] In step four, an improved multiple objective particle swarm optimization (IMOPSO) algorithm is used for solving the problem. Considering the multi-objective, multi-constraint, and mixed-integer nonlinear characteristics of the optimization model, the following improvements are introduced based on the standard IMOPSO algorithm:
[0157] The adaptive inertia weight mechanism uses a biphasic linear decay mechanism to adjust the inertia weight. It maintains a large inertia weight in the early stage of the search to enhance the global exploration capability, and reduces the inertia weight in the later stage to improve the local convergence accuracy. The adaptive parameter control can effectively enhance the algorithm's exploration-exploitation balance capability.
[0158] The strategy of reusing historical experience in elite solutions establishes an elite solution archive, recording the Pareto optimal solutions and their neighborhood information discovered during the search process. This guides particles to gather in the high-quality solution region in subsequent iterations, accelerating convergence. Furthermore, the historical results of GSRI pre-screening are also stored in the elite solution archive: for solutions that pass the pre-screening and have good inner-layer running verification, the sampling probability of their neighborhood particles is increased; for solutions that fail the pre-screening or are on the verge of failure, their violation of short-circuit ratio level and operating condition label are recorded to correct the search direction of subsequent particles.
[0159] The dual-mode oscillation perturbation mechanism applies perturbation to some particles when the population diversity index is below the threshold, in order to increase population diversity and avoid getting trapped in local optima. This perturbation is linked to the critical interval determination of GSRI pre-screening: when a large number of particles are concentrated in the critical interval of GSRI, the perturbation amplitude is increased to accelerate the escape from the critical infeasible region.
[0160] The constraint dominance principle handles multiple constraints, prioritizing feasible solutions over infeasible ones. Feasible solutions are compared based on Pareto dominance, while infeasible solutions are ranked by constraint violation degree. The constraint violation degree is a weighted sum of power balance deviation, MRSCR exceedance degree, energy storage SOC exceedance degree, and grid-connection ratio exceedance degree. The weights can be set according to engineering requirements.
[0161] A two-layer optimization architecture is adopted to decouple capacity configuration and runtime scheduling:
[0162] The outer layer is the capacity configuration layer, using the ratio of wind, solar, thermal, and energy storage capacity and the root / structure network ratio coefficient as optimization variables, and the comprehensive objective function as the fitness function. The IMOPSO algorithm is used for global search. The position vector of each particle is encoded as follows: All the above variables are encoded in real numbers, among which After being normalized to the [0,1] interval using MW or MWh units, it is used in velocity updates. and Directly normalize to the [0,1] interval. Particle velocity updates use the standard MOPSO velocity-position update formula, and out-of-bounds components are handled by boundary absorption or reflection.
[0163] The inner layer is the operation verification layer. For each candidate configuration scheme, the mixed integer linear programming (MILP) model is used in various typical scenarios to perform operation verification with the goal of minimizing the operating cost. The optimal scheduling scheme in each scenario is quickly solved. The solution results of the inner layer (operation cost, wind and solar curtailment, short-circuit margin, etc.) are fed back to the outer layer as the basis for fitness evaluation.
[0164] During the inner-layer operational verification process, the GSRI index is first used for rapid stability pre-screening. Only schemes that fail the pre-screening are subjected to full mixed-integer linear programming (MILP) solution and time-domain simulation verification, forming a hierarchical verification structure of "pre-screening - operational verification - critical scenario time-domain simulation". The full time-domain simulation is concentrated on a few critical schemes and the final optimal scheme, thereby significantly reducing the amount of computation while maintaining the accuracy of stability verification.
[0165] The key parameters of the IMOPSO algorithm were systematically tested and optimized: (1) Population size: test range 50~200, recommended value 100; (2) Maximum number of iterations: test range 200~1000, recommended value 500; (3) Inertia weight range: test range [0.4, 0.9]~[0.2, 1.0], recommended value [0.3, 0.9]; (4) Learning factors c1, c2: test range [1.0, 2.5], recommended value (5) Archive capacity: test range 50~200, recommended value 100. The orthogonal experimental design method is used to test each parameter combination, and the convergence speed, Pareto front distribution uniformity and hypervolume index are used as evaluation criteria to determine the optimal parameter combination.
[0166] After obtaining the Pareto optimal solution set through the algorithm, the AHP-TOPSIS evaluation method is used to select the best compromise solution. This AHP-TOPSIS evaluation method is a multi-index decision-making method combining the Analytic Hierarchy Process (AHP) and the Topology-based Ranking Method (TOPSIS). The decision-maker determines the preference weights for each objective based on the actual engineering needs, calculates the proximity of each Pareto solution to the positive and negative ideal solutions, and selects the solution with the highest proximity as the final recommended solution. Specifically, the AHP judgment matrix is constructed using a 1-9 scale and undergoes a consistency check, requiring a consistency ratio (CR) < 0.1; if CR ≥ 0.1, the judgment matrix is adjusted until the requirement is met. To further reduce subjectivity, the AHP weights and entropy weights can be combined in a ratio of 0.5:0.5 or 0.6:0.4 to form a comprehensive weight before calculating the TOPSIS proximity.
[0167] The final output includes the following configuration schemes: the optimal installed capacity and ratio of wind power, photovoltaic power, thermal power, and energy storage; the optimal ratio coefficient of grid-connected / grid-connected converters for each new energy power station; the power energy ratio and charging / discharging strategy of the energy storage system; system operation indicators under various typical scenarios (new energy utilization rate, short-circuit ratio margin, external power transmission, etc.); and the overall framework design scheme, including power supply layout, voltage support equipment configuration, and external transmission channel utilization scheme.
[0168] The following verification examples illustrate the effectiveness of the present invention.
[0169] Verification Example 1
[0170] Take a large-scale new energy base in Northwest my country as an example. The base is planned to have a wind power capacity of 3,000 MW, a photovoltaic capacity of 2,500 MW, and supporting thermal power capacity of 1,200 MW and electrochemical energy storage of 600 MW / 2,400 MWh. The power will be transmitted to other regions through a ±800 kV ultra-high voltage direct current channel with a transmission capacity of 8,000 MW.
[0171] In terms of scenario set construction, wind speed and irradiance data for the region over the past three years were collected. Combined with the power grid operation boundary conditions, 2,000 initial scenarios were generated. After Kantorovich distance reduction, 30 typical scenarios were formed, covering three short-circuit ratio levels: strong power grid (SCR=3.5), weak power grid (SCR=2.3), and extremely weak power grid (SCR=1.5).
[0172] Regarding the optimized configuration results, after solving the problem for 500 generations using an improved multi-objective particle swarm optimization algorithm, the Pareto optimal solution set was obtained. Through AHP-TOPSIS optimization, the recommended scheme is: 3200MW wind power capacity, 2800MW photovoltaic capacity, 1000MW thermal power capacity, 720MW / 2880MWh energy storage capacity, and a grid-connected converter ratio coefficient α. GFM The value is 0.35 for extremely weak power grid scenarios, 0.20 for weak power grid scenarios, and 0.10 for strong power grid scenarios.
[0173] Among them, the GSRI index is calculated as follows: , ,Right now The equivalent power factor of the new energy power station is taken as 0.98. The GSRI calculations for each scenario are as follows: Strong grid GSRI = 3.5 + 0.10 × 1.62 = 3.66, Weak grid GSRI = 2.3 + 0.20 × 1.62 = 2.62, and Extremely weak grid GSRI = 1.5 + 0.35 × 1.62 = 2.07. Based on the calibrated thresholds of 3.6 for strong grid, 2.6 for weak grid, and 2.0 for extremely weak grid, all three scenarios meet the pre-screening requirements.
[0174] In terms of effectiveness verification, compared with the initial planning scheme, the recommended scheme increased the renewable energy utilization rate from 87.2% to 94.5%, an increase of approximately 7.3 percentage points, reduced the overall investment and operating costs by approximately 10.3%, and shortened the transient voltage recovery time by 42% in extremely weak power grid scenarios. GSRI pre-screening comparison results show that without pre-screening, time-domain simulations need to be performed on all candidate schemes; with pre-screening, approximately 65% of candidate schemes are quickly eliminated, with a pre-screening accuracy of approximately 92%, a recall rate of approximately 96%, and an overall solution time reduction of approximately 60%.
[0175] Verification Example 2
[0176] This example analyzes the impact of the mesh / network ratio on system stability. Under the same site scenario, α is set separately. GFM =0 (pure follow-up network), α GFM =0.15、α GFM =0.30、α GFM A comparative analysis of four configuration schemes with a value of 0.50 is conducted.
[0177] The results show that the pure grid-following scheme exhibits significant transient voltage instability in extremely weak grid scenarios with SCR < 2.0, with a phase-locked loop (PLL) unlock probability of up to 25%; α GFM When α = 0.15, the system can barely maintain stability under extremely weak power grid scenarios, but the margin is insufficient; GFM When α = 0.30, the system meets the stability requirements in all scenarios, and the utilization rate of new energy reaches 93.8%; GFM Stability is further improved when the value is 0.50, but the investment cost increases by about 18%.
[0178] Considering both economic efficiency and stability, α in extremely weak power grid scenarios GFM =0.30~0.35 is the optimal ratio range. Verification using the GSRI index: In extremely weak power grid scenarios, α GFM When α = 0.30, GSRI = 1.5 + 0.30 × 1.62 = 1.99, slightly lower than the threshold of 2.0, and within the critical range. Detailed time-domain simulation verification confirms that the stability requirements are met. GFM When α = 0.35, GSRI = 1.5 + 0.35 × 1.62 = 2.07, which is greater than the threshold of 2.0, thus passing the pre-screening. Further, based on the pure tracking network scheme, α... GFM When α = 0.30, the dynamic voltage support capability index is improved by about 35%, and the transient voltage recovery time in extremely weak power grid scenarios is shortened by about 42%; GFM When the value is 0.50, the dynamic voltage support capability index is improved by about 45%, but the investment cost increases by about 18%.
[0179] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any changes made based on the design principles of the present invention, or any non-creative modifications made thereon, shall fall within the scope of protection of the present invention.
Claims
1. A method for optimizing the configuration of heterogeneous power sources in a new energy base to adapt to multiple operating conditions, characterized in that, Includes the following steps: Step 1: Construct a set of typical scenarios with multiple operating conditions: Collect wind and solar power output data and grid operation boundary conditions from new energy bases, and construct a set of typical power generation output time series scenarios covering different short-circuit ratio levels and multiple operating condition combinations; Step 2: Establish a heterogeneous power source optimization configuration model: Using the installed capacity of wind power, photovoltaic power, thermal power, and energy storage, the energy storage power ratio, and the matching coefficient of grid-connected and grid-connected converters of each new energy power station as decision variables, and taking comprehensive investment and operating costs, dynamic voltage support capacity, and new energy utilization rate as multiple objectives, and setting power balance constraints, multi-station short-circuit ratio constraints, conventional unit operation constraints, energy storage operation constraints, DC transmission channel constraints, and grid-connected / grid-connected matching constraints, an optimization configuration model is established. Step 3: Rapid stability verification: Introduce a ratio index based on power grid strength assessment to perform rapid stability pre-screening of candidate configuration schemes, and only perform time-domain simulation verification on candidate configuration schemes that are in the critical range of the pre-screening. Step 4: Optimization and Solution: An improved multi-objective particle swarm optimization algorithm is used to solve the optimization configuration model with power capacity configuration as the outer layer and operation verification as the inner layer, so as to obtain the optimal power configuration scheme.
2. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 1, characterized in that, The process of step one includes: Collect meteorological data on wind speed, solar irradiance, and temperature, as well as historical power output data from wind and solar power plants, and collect grid operation boundary conditions; Based on the short-circuit ratio at the grid connection point, the power grid strength is divided into three levels: strong grid, weak grid, and extremely weak grid. Combined with load level, transmission channel operation mode, and conventional unit start-up mode, a three-dimensional operating condition combination matrix of short-circuit ratio level × load level × transmission mode is constructed.
3. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 2, characterized in that, In step one, considering the spatiotemporal correlation between wind power output and photovoltaic power output, a joint probability distribution model of wind power output and photovoltaic power output is constructed by using a Copula function, and a large-scale initial scene set is generated by Latin hypercube sampling; the initial scene set is simplified by a scene reduction method based on Kantorovich distance to form the typical scene set.
4. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 3, characterized in that, The scenario reduction adopts the synchronous back-substitution elimination method to gradually eliminate the scenario pairs with the smallest probability distance and merge their probabilities into the retained scenarios; the typical scenario set formed has a scale of 20 to 50 scenarios, each scenario contains the wind and solar power output time series curves on a time scale of 24 hours or longer, and is labeled with the corresponding short-circuit ratio level and operating condition label.
5. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 1, characterized in that, In step two, the energy storage power-to-energy ratio is defined as the ratio of energy storage power capacity to energy storage capacity; the energy storage duration is defined as the ratio of energy storage capacity to energy storage power capacity.
6. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 5, characterized in that, In step two, the matching ratio coefficient between grid-type and grid-structured converters is defined as the proportion of the capacity of grid-structured converters to the total converter capacity of the station, with a value range of 0 to 1.
7. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 6, characterized in that, In step two, the multi-objective includes three sub-objectives: minimizing the overall investment and operating cost, maximizing the dynamic voltage support capacity, and maximizing the utilization rate of new energy sources. The three sub-objectives are normalized and then weighted and summed using the weight coefficient method to transform them into a comprehensive objective function. Among them, the dynamic voltage support capacity is represented by the sum of the dynamic voltage support capacity indicators of the system under each scenario according to the scenario probability weight. The new energy utilization rate is represented by the ratio of the sum of the actual wind power and photovoltaic power consumed in each time period under each scenario to the sum of the theoretical available power.
8. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 7, characterized in that, In step two, the short-circuit ratio constraint for multiple power stations is determined by the short-circuit capacity of the power station, the active power output of the new energy power station, and the mutual impedance and self-impedance of the system impedance matrix, and is not less than the minimum short-circuit ratio threshold. After configuring grid-type converters or distributed synchronous condensers, the short-circuit ratio of the multiple power stations is increased by increasing the system short-circuit capacity.
9. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions according to any one of claims 1 to 8, characterized in that, In step three, the matching ratio index is determined by the sum of the basic short-circuit ratio and the equivalent short-circuit ratio increment of the grid-type converter. The basic short-circuit ratio is the ratio of the system's basic short-circuit capacity excluding the contribution of the grid-type converter to the total apparent power of new energy sources. The equivalent short-circuit ratio increment is determined by the grid-type converter matching ratio coefficient and the equivalent enhancement coefficient of the short-circuit ratio per unit capacity of the grid-type converter. The matching ratio index increases with the increase of the grid-type converter matching ratio coefficient. When the matching ratio index of a candidate configuration scheme is greater than the preset threshold of the corresponding short-circuit ratio level, the candidate configuration scheme is determined to meet the stability requirements. When the matching ratio index is within a preset critical range, time-domain simulation verification is triggered. When the matching ratio index is lower than the preset lower limit of the critical range, the candidate configuration scheme is determined to not meet the stability requirements and is directly eliminated.
10. The method for optimizing the configuration of heterogeneous power sources in a new energy base adaptable to multiple operating conditions as described in claim 9, characterized in that, In step four, the improved multi-objective particle swarm optimization algorithm introduces an adaptive inertial weight mechanism, an elite historical experience reuse strategy, a dual-mode oscillation disturbance mechanism, and a constraint domination principle on the basis of the standard multi-objective particle swarm optimization algorithm. The outer layer uses the ratio of wind, solar, thermal and energy storage capacity and the ratio of root / structure network as optimization variables to perform a global search. The inner layer uses a mixed integer linear programming model to perform operation verification in various typical scenarios, and the operation verification results of the inner layer are fed back to the outer layer as the basis for fitness evaluation. After obtaining the Pareto optimal solution set, a multi-criteria decision-making method is used to select the best compromise solution.
