Offshore wind power flexible direct system converter station and fan converter control parameter collaborative optimization method and system

Abstract

This invention relates to the field of flexible DC transmission technology for offshore wind power, and discloses a method and system for collaborative optimization of control parameters of converter stations and wind turbine converters in offshore wind power flexible DC systems. The method includes: constructing an electromagnetic transient simulation model comprising a sending-end converter station, a receiving-end converter station, a DC transmission line, and a wind farm; introducing multiple disturbance conditions and combining variations in control parameters, recording dynamic response process data; establishing a collaborative optimization criterion composed of multiple performance indicators directly mapped from the dynamic response process data; and employing a multi-objective evolutionary optimization mechanism with control parameters as the variables to be optimized and performance indicators as the optimization guide for iterative search. This invention achieves global collaborative configuration of control parameters for converter stations and wind turbine converters, solving the problem of traditional step-by-step tuning methods neglecting dynamic interaction effects, and significantly improving the overall operational robustness of offshore wind power flexible DC systems under multiple disturbance conditions.

Description

Technical Field

[0001] This invention relates to the field of flexible DC transmission technology for offshore wind power, specifically to a method and system for collaborative optimization of control parameters of converter stations and wind turbine converters in offshore wind power flexible DC transmission systems. Background Technology

[0002] As the large-scale development of offshore wind power extends to deep-sea areas, flexible DC transmission technology has become the mainstream solution for long-distance, large-capacity offshore wind power transmission. In the offshore wind power flexible DC system, the sending-end converter station is responsible for converting the unstable AC power collected by the wind farm into DC power, the receiving-end converter station is responsible for inverting the DC power and feeding it into the onshore power grid, and the wind turbine converter controls the power output and grid-connected operation characteristics of the wind turbine. The control parameters of these three are coupled with each other. Improper parameter settings in any link will affect the dynamic and stable operation of the entire system.

[0003] In existing technologies, the tuning of converter station control parameters and wind turbine converter control parameters is typically performed independently in steps. Designers first tune the control parameters of the sending-end and receiving-end converter stations based on the dynamic response requirements of the converter station itself, then tune the control parameters of the wind turbine converter according to the grid connection guidelines for wind turbines, and finally verify the overall system performance through simulation. This serial tuning method fails to fully consider the dynamic interaction between the converter station and the wind turbine converter, leading to problems such as slow voltage recovery, increased power oscillations, and even wind turbine disconnection when dealing with complex operating conditions such as AC-side faults and DC-side disturbances. Especially in scenarios with multiple stations clustered together, the mismatch between the converter station and wind turbine converter control parameters further amplifies the transient overshoot and oscillation risks of the system, seriously threatening the safe and stable operation of offshore wind power flexible DC systems. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for co-optimizing the control parameters of the converter station and the wind turbine converter in an offshore wind power flexible DC system, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for collaborative optimization of control parameters of the converter station and wind turbine converter in an offshore wind power flexible DC system, comprising the following steps: S1. Construct an electromagnetic transient simulation model of an offshore wind power flexible DC system. The electromagnetic transient simulation model shall include at least the equivalent circuit and control system models of the sending-end converter station, receiving-end converter station, DC transmission line, and wind farm. S2. Based on the electromagnetic transient simulation model, a variety of preset disturbance conditions are introduced, and the control parameters of the sending-end converter station and the wind turbine converter are combined and varied. The dynamic response process data of the key electrical quantities of the sending-end converter station and the wind turbine converter under different parameter combinations and disturbance conditions are recorded. S3. Establish a set of collaborative optimization criteria. The collaborative optimization criteria include multiple performance indicators for evaluating the quality of dynamic response process data, and each performance indicator is directly mapped from the dynamic response process data. S4. Using a multi-objective evolutionary optimization mechanism, the control parameters of the sending-end converter station and the control parameters of the wind turbine converter are the variables to be optimized. Multiple performance indicators in the collaborative optimization criteria are used as optimization guides to repeatedly call and iteratively search the electromagnetic transient simulation model. S5. From the numerous parameter combinations obtained through iterative search, identify a set of control parameters for the sending-end converter station and the wind turbine converter that achieve the optimal balance of multiple performance indicators as a whole, and output them as the result of collaborative optimization.

[0006] As a preferred embodiment of the present invention, the step S2 of recording dynamic response process data specifically includes: Based on the location and type of disturbance, the dynamic response process data is divided into the sending-end AC side disturbance response dataset, the DC side disturbance response dataset, and the receiving-end AC side disturbance response dataset. In the disturbance response dataset of the sending-end AC side, extract the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending-end converter station; In the DC-side disturbance response dataset, extract the transient overshoot peak data and the time constant data of fluctuation decay of the DC voltage at the receiving-end converter station; In the AC-side disturbance response dataset at the receiving end, the transient drop duration data of the wind turbine grid connection point voltage and the transient fluctuation energy data of the wind farm output power are extracted.

[0007] As a preferred embodiment of the present invention, the establishment of the collaborative optimization criterion in step S3 specifically includes: Based on the transient drop envelope data and recovery trajectory oscillation mode data of the AC bus voltage of the sending-end converter station, a first performance index group is constructed. The first performance index group is used to characterize the voltage support strength and recovery stability of the sending-end converter station after disturbance. Based on the transient overshoot peak data and fluctuation decay time constant data of the DC voltage at the receiving-end converter station, a second performance index group is constructed. The second performance index group is used to characterize the voltage robustness and dynamic response speed of the DC grid when facing power surges. Based on the transient drop duration data of the grid connection point voltage of the wind turbine and the transient fluctuation energy data of the output power of the wind farm, a third performance index group is constructed. The third performance index group is used to characterize the wind farm's ability to operate without disconnecting from the grid and the stability of its power output during grid disturbances. The first, second, and third performance index groups are jointly established as the collaborative optimization criteria.

[0008] As a preferred embodiment of the present invention, step S4 employs a multi-objective evolutionary optimization mechanism for iterative search, specifically including: Initialize a search population containing multiple sets of control parameters for the sending-end converter station and control parameters for the wind turbine converter, with each set of parameters as an independent search entity; Each individual in the search population is assigned to an electromagnetic transient simulation model in turn, and multiple preset disturbance conditions are simulated to obtain the actual performance values ​​of the first performance index group, the second performance index group and the third performance index group corresponding to each individual. Based on the actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy, and the sparseness of the distribution of individuals in the target space is evaluated by the crowding distance within the same level. Based on the Pareto front level and the crowding distance, select some individuals from the current population for crossover and mutation operations to generate a new generation of search population, and repeat the above process until the preset iteration termination condition is met.

[0009] As a preferred embodiment of the present invention, step S5, which involves identifying a set of parameter combinations that achieve an optimal balance among multiple performance indicators, specifically includes: From the final generation of search population that meets the iteration termination condition, select all search individuals located at the first Pareto front level to form a candidate parameter solution set; In the candidate parameter solution set, the parameter individual with the minimum degree of conflict among the first performance index group, the second performance index group and the third performance index group is identified. The minimum degree of conflict is measured by the comprehensive deviation between the actual performance value and the ideal performance value of the specific index within each performance index group. The control parameters of the sending-end converter station and the control parameters of the wind turbine converter corresponding to the identified individual parameters are determined as the final collaborative optimization result.

[0010] A system for collaborative optimization of control parameters of offshore wind power flexible DC system converter station and wind turbine converter for implementing any of the methods described above includes: The model building unit builds an electromagnetic transient simulation model that includes the sending-end converter station, receiving-end converter station, DC transmission line and wind farm equivalent model, based on the actual topology and component characteristics of the offshore wind power flexible DC system. The data recording unit connects to the electromagnetic transient simulation model of the model building unit. Based on various preset disturbance conditions, it combines and changes the control parameters of the sending-end converter station and the wind turbine converter in the model, and records the dynamic response process data of the key electrical quantities of the sending-end converter station and the wind turbine converter under different parameter combinations. The criteria construction unit receives dynamic response process data recorded by the data recording unit and establishes a set of collaborative optimization criteria that includes multiple performance indicators, all of which are directly mapped from the dynamic response process data. The iterative search unit receives the collaborative optimization criteria from the criterion construction unit, adopts a multi-objective evolutionary optimization mechanism, takes the control parameters of the sending-end converter station and the control parameters of the wind turbine converter as variables to be optimized, uses multiple performance indicators in the collaborative optimization criteria as optimization guides, and repeatedly calls the electromagnetic transient simulation model of the model building unit to perform iterative optimization. The result output unit receives the optimization results from the iterative search unit, and identifies a set of control parameters for the sending-end converter station and the wind turbine converter that achieves the optimal balance of multiple performance indicators from the final parameter combination. This set is then output as a collaborative optimization configuration scheme.

[0011] As a preferred technical solution of the present invention, when the data recording unit records dynamic response process data, it divides the dynamic response process data into sending-end AC side disturbance response dataset, DC side disturbance response dataset and receiving-end AC side disturbance response dataset according to the disturbance occurrence location and disturbance type. In the AC side disturbance response dataset of the sending end, the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending end converter station are extracted; in the DC side disturbance response dataset, the transient overshoot peak data and the time constant data of fluctuation decay of the DC voltage of the receiving end converter station are extracted. In the AC-side disturbance response dataset at the receiving end, the transient drop duration data of the wind turbine grid connection point voltage and the transient fluctuation energy data of the wind farm output power are extracted.

[0012] As a preferred technical solution of the present invention, when the criterion construction unit establishes the collaborative optimization criterion, it constructs a first set of performance indicators to characterize the voltage support strength and recovery stability of the sending-end converter station based on the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending-end converter station. Based on the transient overshoot peak data and fluctuation decay time constant data of the DC voltage at the receiving-end converter station, a second set of performance indicators is constructed to characterize the voltage robustness and dynamic response speed of the DC grid. Based on the transient drop duration data of the grid connection voltage of wind turbines and the transient fluctuation energy data of the output power of wind farms, a third set of performance indicators is constructed to characterize the ability of wind farms to operate without disconnecting from the grid and the stability of power output. The first, second, and third performance index groups are jointly established as the collaborative optimization criteria.

[0013] As a preferred technical solution of the present invention, when the iterative search unit performs iterative search using a multi-objective evolutionary optimization mechanism, it initializes a search population containing multiple sets of sending-end converter station control parameters and wind turbine converter control parameters. Each search individual is sequentially assigned to an electromagnetic transient simulation model and the simulation is executed to obtain the actual performance values ​​of the first performance index group, the second performance index group and the third performance index group corresponding to each search individual. Based on actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy, and the crowding distance of individuals is evaluated within the same level. Based on the Pareto front hierarchy and the distance between crowding levels, select individuals are crossovered and mutated to generate a new generation of search population until the preset iteration termination condition is met.

[0014] As a preferred technical solution of the present invention, when the result output unit identifies parameter combinations, it selects all search individuals located at the first Pareto front level from the final generation of search population that meets the iteration termination condition to form a candidate parameter solution set. In the candidate parameter solution set, the parameter individual with the minimum degree of conflict among the first performance index group, the second performance index group and the third performance index group is identified. The minimum degree of conflict is measured by the comprehensive deviation between the actual performance value and the ideal performance value of the specific index within each performance index group. The control parameters of the sending-end converter station and the control parameters of the wind turbine converter corresponding to the identified individual parameters are determined as the final collaborative optimization result.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention constructs an electromagnetic transient simulation model that includes an sending-end converter station, a receiving-end converter station, a DC transmission line, and an equivalent circuit of a wind farm. In this model, various disturbance conditions are introduced to combine and change the control parameters of the converter station and the wind turbine converter, thereby achieving a comprehensive characterization of the dynamic interaction process between the converter station and the wind turbine converter, laying a solid data foundation for subsequent collaborative optimization.

[0016] 2. This invention directly maps dynamic response process data into a set of performance indicators with clear physical meaning, and constructs a multi-dimensional collaborative optimization criterion that includes sending-end voltage support capability, DC voltage stability capability, and wind turbine grid-friendliness. This avoids the drawbacks of traditional methods that rely on empirical weight assignment, and makes the optimization orientation highly consistent with the actual operating requirements of the system.

[0017] 3. This invention employs a multi-objective evolutionary optimization mechanism, using the control parameters of the sending-end converter station and the control parameters of the wind turbine converter as variables to be optimized, and using multiple performance indicators in the collaborative optimization criteria as optimization guides for iterative search. It can automatically search for a set of balanced collaborative solutions among multiple conflicting optimization objectives such as the AC voltage recovery speed of the sending end, the DC voltage fluctuation suppression capability, and the low voltage ride-through performance of the wind turbine, thus solving the problem of traditional single-objective optimization methods that fail to address all aspects.

[0018] 4. This invention identifies a set of parameter combinations from numerous parameter combinations obtained through iterative search that achieves an optimal balance among multiple performance indicators. It realizes the global coordinated configuration of control parameters between the converter station and the wind turbine converter, enabling the system to exhibit excellent dynamic response characteristics when facing different types of disturbances, and significantly improving the overall operational robustness and reliability of the offshore wind power flexible DC system.

[0019] 5. The method and system provided by this invention have clear physical logic and operable technical path, and can be directly applied to the control parameter tuning work of actual offshore wind power flexible DC engineering, providing scientific and efficient decision support for engineering designers, and have good engineering promotion value. Attached Figure Description

[0020] Figure 1 This is an overall flowchart of the method for coordinated optimization of control parameters of converter station and wind turbine converter in offshore wind power flexible DC system according to the present invention; Figure 2 This is a structural framework diagram of the system for collaborative optimization of control parameters of the converter station and wind turbine converter in the offshore wind power flexible DC system of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1

[0023] A method for collaborative optimization of control parameters between the converter station and the wind turbine converter in an offshore wind power flexible DC system includes the following steps: S1. Construct an electromagnetic transient simulation model of an offshore wind power flexible DC system. The electromagnetic transient simulation model shall include at least the equivalent circuit and control system models of the sending-end converter station, receiving-end converter station, DC transmission line, and wind farm. S2. Based on the electromagnetic transient simulation model, a variety of preset disturbance conditions are introduced, and the control parameters of the sending-end converter station and the wind turbine converter are combined and varied. The dynamic response process data of the key electrical quantities of the sending-end converter station and the wind turbine converter under different parameter combinations and disturbance conditions are recorded. S3. Establish a set of collaborative optimization criteria. The collaborative optimization criteria include multiple performance indicators for evaluating the quality of dynamic response process data, and each performance indicator is directly mapped from the dynamic response process data. S4. Using a multi-objective evolutionary optimization mechanism, the control parameters of the sending-end converter station and the control parameters of the wind turbine converter are the variables to be optimized. Multiple performance indicators in the collaborative optimization criteria are used as optimization guides to repeatedly call and iteratively search the electromagnetic transient simulation model. S5. From the numerous parameter combinations obtained through iterative search, identify a set of control parameters for the sending-end converter station and the wind turbine converter that achieve the optimal balance of multiple performance indicators as a whole, and output them as the result of collaborative optimization.

[0024] Furthermore, step S2 records dynamic response process data, specifically including: Based on the location and type of disturbance, the dynamic response process data is divided into the sending-end AC side disturbance response dataset, the DC side disturbance response dataset, and the receiving-end AC side disturbance response dataset. In the disturbance response dataset of the sending-end AC side, extract the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending-end converter station; In the DC-side disturbance response dataset, extract the transient overshoot peak data and the time constant data of fluctuation decay of the DC voltage at the receiving-end converter station; In the AC-side disturbance response dataset at the receiving end, the transient drop duration data of the wind turbine grid connection point voltage and the transient fluctuation energy data of the wind farm output power are extracted.

[0025] Furthermore, step S3 establishes a collaborative optimization criterion, which specifically includes: Based on the transient drop envelope data and recovery trajectory oscillation mode data of the AC bus voltage of the sending-end converter station, a first performance index group is constructed. The first performance index group is used to characterize the voltage support strength and recovery stability of the sending-end converter station after disturbance. Based on the transient overshoot peak data and fluctuation decay time constant data of the DC voltage at the receiving-end converter station, a second performance index group is constructed. The second performance index group is used to characterize the voltage robustness and dynamic response speed of the DC grid when facing power surges. Based on the transient drop duration data of the grid connection point voltage of the wind turbine and the transient fluctuation energy data of the output power of the wind farm, a third performance index group is constructed. The third performance index group is used to characterize the wind farm's ability to operate without disconnecting from the grid and the stability of its power output during grid disturbances. The first, second, and third performance index groups are jointly established as the collaborative optimization criteria.

[0026] Furthermore, step S4 employs a multi-objective evolutionary optimization mechanism for iterative search, specifically including: Initialize a search population containing multiple sets of control parameters for the sending-end converter station and control parameters for the wind turbine converter, with each set of parameters as an independent search entity; Each individual in the search population is assigned to an electromagnetic transient simulation model in turn, and multiple preset disturbance conditions are simulated to obtain the actual performance values ​​of the first performance index group, the second performance index group and the third performance index group corresponding to each individual. Based on the actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy, and the sparseness of the distribution of individuals in the target space is evaluated by the crowding distance within the same level. Based on the Pareto front level and the crowding distance, select some individuals from the current population for crossover and mutation operations to generate a new generation of search population, and repeat the above process until the preset iteration termination condition is met.

[0027] Furthermore, step S5 involves identifying a set of parameter combinations that achieve the optimal balance among multiple performance indicators, specifically including: From the final generation of search population that meets the iteration termination condition, select all search individuals located at the first Pareto front level to form a candidate parameter solution set; In the candidate parameter solution set, the parameter individual with the minimum degree of conflict among the first performance index group, the second performance index group and the third performance index group is identified. The minimum degree of conflict is measured by the comprehensive deviation between the actual performance value and the ideal performance value of the specific index within each performance index group. The control parameters of the sending-end converter station and the control parameters of the wind turbine converter corresponding to the identified individual parameters are determined as the final collaborative optimization result.

[0028] A system for collaborative optimization of control parameters of offshore wind power flexible DC system converter station and wind turbine converter, used to implement any of the above methods, includes: The model building unit builds an electromagnetic transient simulation model that includes the sending-end converter station, receiving-end converter station, DC transmission line and wind farm equivalent model, based on the actual topology and component characteristics of the offshore wind power flexible DC system. The data recording unit connects to the electromagnetic transient simulation model of the model building unit. Based on various preset disturbance conditions, it combines and changes the control parameters of the sending-end converter station and the wind turbine converter in the model, and records the dynamic response process data of the key electrical quantities of the sending-end converter station and the wind turbine converter under different parameter combinations. The criteria construction unit receives dynamic response process data recorded by the data recording unit and establishes a set of collaborative optimization criteria that includes multiple performance indicators, all of which are directly mapped from the dynamic response process data. The iterative search unit receives the collaborative optimization criteria from the criterion construction unit, adopts a multi-objective evolutionary optimization mechanism, takes the control parameters of the sending-end converter station and the control parameters of the wind turbine converter as variables to be optimized, uses multiple performance indicators in the collaborative optimization criteria as optimization guides, and repeatedly calls the electromagnetic transient simulation model of the model building unit to perform iterative optimization. The result output unit receives the optimization results from the iterative search unit, and identifies a set of control parameters for the sending-end converter station and the wind turbine converter that achieves the optimal balance of multiple performance indicators from the final parameter combination. This set is then output as a collaborative optimization configuration scheme.

[0029] Furthermore, when the data recording unit records dynamic response process data, it divides the dynamic response process data into sending-end AC side disturbance response dataset, DC side disturbance response dataset, and receiving-end AC side disturbance response dataset according to the location and type of disturbance. In the AC side disturbance response dataset of the sending end, the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending end converter station are extracted; in the DC side disturbance response dataset, the transient overshoot peak data and the time constant data of fluctuation decay of the DC voltage of the receiving end converter station are extracted. In the AC-side disturbance response dataset at the receiving end, the transient drop duration data of the wind turbine grid connection point voltage and the transient fluctuation energy data of the wind farm output power are extracted.

[0030] Furthermore, when establishing collaborative optimization criteria, the criterion construction unit constructs a first set of performance indicators to characterize the voltage support strength and recovery stability of the sending-end converter station based on the transient drop envelope data of the AC bus voltage at the sending-end converter station and the oscillation mode data of the recovery trajectory. Based on the transient overshoot peak data and fluctuation decay time constant data of the DC voltage at the receiving-end converter station, a second set of performance indicators is constructed to characterize the voltage robustness and dynamic response speed of the DC grid. Based on the transient drop duration data of the grid connection voltage of wind turbines and the transient fluctuation energy data of the output power of wind farms, a third set of performance indicators is constructed to characterize the ability of wind farms to operate without disconnecting from the grid and the stability of power output. The first, second, and third performance index groups are jointly established as the collaborative optimization criteria.

[0031] Furthermore, when the iterative search unit performs iterative search using a multi-objective evolutionary optimization mechanism, it initializes a search population containing multiple sets of control parameters for the sending-end converter station and control parameters for the wind turbine converter. Each search individual is sequentially assigned to an electromagnetic transient simulation model and the simulation is executed to obtain the actual performance values ​​of the first performance index group, the second performance index group and the third performance index group corresponding to each search individual. Based on actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy, and the crowding distance of individuals is evaluated within the same level. Based on the Pareto front hierarchy and the distance between crowding levels, select individuals are crossovered and mutated to generate a new generation of search population until the preset iteration termination condition is met.

[0032] Furthermore, when the result output unit identifies parameter combinations, it selects all search individuals located at the first Pareto front level from the final generation of search population that meets the iteration termination condition to form a candidate parameter solution set. In the candidate parameter solution set, the parameter individual with the minimum degree of conflict among the first performance index group, the second performance index group and the third performance index group is identified. The minimum degree of conflict is measured by the comprehensive deviation between the actual performance value and the ideal performance value of the specific index within each performance index group. The control parameters of the sending-end converter station and the control parameters of the wind turbine converter corresponding to the identified individual parameters are determined as the final collaborative optimization result.

[0033] Example 2

[0034] This embodiment takes the Rudong offshore wind power flexible DC transmission project in Jiangsu Province as an example to further illustrate the specific implementation process of the collaborative optimization method of control parameters of the converter station and wind turbine converter in the offshore wind power flexible DC system.

[0035] In step S1, an electromagnetic transient simulation model of the offshore wind power flexible DC system is constructed. Based on the actual parameters of the Rudong project, the sending-end converter station adopts a modular multilevel converter topology with a rated capacity of 1100MW and a DC voltage of ±320kV. The receiving-end converter station is located on land and also adopts a modular multilevel converter topology. The DC transmission line uses a 100km long DC cable. The wind farm aggregation model consists of 50 permanent magnet direct-drive wind turbines with a single unit capacity of 5MW. Each wind turbine is equipped with a full-power converter. In the simulation platform, the converter station control system adopts a direct current control strategy, and the wind turbine converter control system adopts a grid voltage-oriented vector control strategy. The control system model includes core control components such as the inner current loop, the outer power loop, and the phase-locked loop.

[0036] In step S2, several preset disturbance conditions are introduced, including a three-phase short-circuit fault on the sending-end AC bus with a duration of 100ms; a single-phase ground fault on the receiving-end AC bus with a duration of 150ms; and a single-pole ground fault on the DC line with a duration of 50ms. The proportional and integral coefficients of the current inner loop and the proportional and integral coefficients of the power outer loop in the control parameters of the sending-end converter station, as well as the proportional and integral coefficients of the current inner loop and the proportional and integral coefficients of the phase-locked loop in the control parameters of the wind turbine converter, are combined and varied. The parameter variation range is set to 0.5 to 2.0 times the rated value. Simulations of the above three disturbance conditions are performed under each parameter combination, and the waveforms of the AC bus voltage of the sending-end converter station, the DC voltage of the receiving-end converter station, the voltage waveform of the wind turbine grid connection point, and the active power output waveform of the wind farm are recorded.

[0037] In step S3, a collaborative optimization criterion is established. Transient sag envelope data and recovery trajectory oscillation mode data of the AC bus voltage at the sending-end converter station are extracted from the recorded waveform data to construct a first performance index set. This index set includes three specific indicators: voltage sag depth, voltage recovery time, and the number of oscillations during the recovery process. These indicators collectively characterize the voltage support strength and recovery stability of the sending-end converter station after a disturbance. Transient overshoot peak data and fluctuation decay time constant data are extracted from the DC voltage waveform at the receiving-end converter station to construct a second performance index set. This index set includes three specific indicators: voltage overshoot amplitude, voltage fluctuation duration, and fluctuation decay rate. The three performance indexes are used to characterize the voltage robustness and dynamic response speed of the DC grid when facing power surges. Transient sag duration data and transient fluctuation energy data are extracted from the voltage waveform at the wind turbine grid connection point and the output power waveform of the wind farm to construct a third performance index group. This index group includes three specific indicators: the length of time the wind turbine remains connected to the grid during the voltage sag, the smoothness of the power recovery after the voltage recovers, and the energy integral value covered by the power fluctuation. Together, they characterize the wind farm's ability to operate without disconnecting from the grid and the stability of its power output during grid disturbances. The first, second, and third performance index groups are jointly established as a collaborative optimization criterion.

[0038] Through the progressive development of the above technical logic, this embodiment can transform complex dynamic response waveforms into a set of performance indicators with clear physical meaning, providing a clear evaluation basis for the subsequent optimization process. The resulting technical effect is that the construction of optimization criteria no longer relies on empirical weight assignment, but comes directly from the dynamic response characteristics of the system itself, making the optimization orientation highly consistent with the actual operating requirements of the system and significantly improving the engineering applicability of the optimization results.

[0039] Example 3

[0040] This embodiment takes the Guangdong Yangjiang offshore wind power flexible direct transmission project as an example to further illustrate the specific implementation process and technical logic of steps S4 and S5 based on embodiment 2.

[0041] In step S4, an iterative search is performed using a multi-objective evolutionary optimization mechanism. A search population containing 100 sets of control parameters for the sending-end converter station and control parameters for the wind turbine converter is initialized. Each set of parameters is treated as an independent search individual. The parameter values ​​are randomly generated within the range of 0.5 to 2.0 times the rated value. Each search individual in the search population is assigned to an electromagnetic transient simulation model in sequence. Simulation is performed according to the three disturbance conditions set in Example 2 to obtain the actual performance values ​​of the first performance index group, the second performance index group, and the third performance index group corresponding to each search individual.

[0042] Based on actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy. The specific operation of non-dominated ranking is as follows: for any two search individuals, if one individual is not inferior to the other individual in all specific metrics of the first, second, and third performance metric groups, and is superior to the other individual in at least one metric, then that individual is considered to dominate the other individual. All search individuals not dominated by any other individual are assigned to the first Pareto front level; search individuals dominated only by individuals in the first level are assigned to the second Pareto front level; and so on, completing the hierarchical division of all individuals. Within the same level, the sparseness of an individual's distribution in the target space is assessed by crowding distance. The larger the crowding distance, the fewer other individuals around that individual, and the stronger its representativeness in the target space. Individuals at the higher levels of the Pareto front are selected first, and individuals with larger crowding distances within the same level are selected first. A selection of individuals from the current population are selected for simulated binary crossover and polynomial mutation operations to generate a new generation of search population containing 100 individuals. The above iterative process of simulation evaluation, non-dominated sorting, crowding evaluation, and selective breeding is repeated until the preset 50-generation iteration termination condition is reached.

[0043] In step S5, a set of parameter combinations that achieves an optimal trade-off among multiple performance indicators is identified. From the final generation of search population that meets the iteration termination condition, all search individuals located at the first Pareto front level are selected to form a candidate parameter solution set. In the candidate parameter solution set, the parameter individual with the minimum conflict degree among the first, second, and third performance indicator groups is identified. The process of identifying the minimum conflict degree is as follows: for the three specific indicators in the first performance indicator group—voltage drop depth, voltage recovery time, and recovery oscillation number—the optimal value that appears in the entire search process is determined as the ideal performance value. Similarly, the ideal performance value is determined for each specific indicator in the second and third performance indicator groups. The Euclidean distance between the actual performance value and the ideal performance value of each candidate parameter individual on each specific indicator is calculated, and the Euclidean distances of all specific indicators are summed to obtain the comprehensive deviation degree of the individual. The parameter individual with the smallest comprehensive deviation degree is selected, and its corresponding sending-end converter station control parameters and wind turbine converter control parameters are determined as the final collaborative optimization result.

[0044] Through rigorous deduction of the above technical logic, this embodiment realizes a closed-loop process from population initialization to Pareto front screening. The resulting technical effect is that the multi-objective evolutionary optimization mechanism can automatically search for a set of balanced cooperative solutions among multiple conflicting optimization objectives such as ensuring the voltage support capability of the converter station, the DC voltage stability capability, and the grid-friendliness of the wind turbine. This avoids the drawbacks of traditional single-objective optimization methods that focus on one aspect while neglecting another. The final output parameter combination enables the system to achieve the optimal trade-off between the AC voltage recovery speed at the sending end, the DC voltage fluctuation suppression capability, and the low voltage ride-through performance of the wind turbine when facing different types of disturbances, thereby comprehensively improving the overall operational robustness of the offshore wind power flexible DC system.

[0045] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for collaborative optimization of control parameters of converter station and wind turbine converter in offshore wind power flexible DC system, characterized in that, Includes the following steps: S1. Construct an electromagnetic transient simulation model of an offshore wind power flexible DC system. The electromagnetic transient simulation model shall include at least the equivalent circuit and control system models of the sending-end converter station, receiving-end converter station, DC transmission line, and wind farm. S2. Based on the electromagnetic transient simulation model, a variety of preset disturbance conditions are introduced, and the control parameters of the sending-end converter station and the wind turbine converter are combined and varied. The dynamic response process data of the key electrical quantities of the sending-end converter station and the wind turbine converter under different parameter combinations and disturbance conditions are recorded. S3. Establish a set of collaborative optimization criteria. The collaborative optimization criteria include multiple performance indicators for evaluating the quality of dynamic response process data, and each performance indicator is directly mapped from the dynamic response process data. S4. Using a multi-objective evolutionary optimization mechanism, the control parameters of the sending-end converter station and the control parameters of the wind turbine converter are the variables to be optimized. Multiple performance indicators in the collaborative optimization criteria are used as optimization guides to repeatedly call and iteratively search the electromagnetic transient simulation model. S5. From the numerous parameter combinations obtained through iterative search, identify a set of control parameters for the sending-end converter station and the wind turbine converter that achieve the optimal balance of multiple performance indicators as a whole, and output them as the result of collaborative optimization.

2. The method according to claim 1, characterized in that, The step S2, which records the dynamic response process data, specifically includes: Based on the location and type of disturbance, the dynamic response process data is divided into the sending-end AC side disturbance response dataset, the DC side disturbance response dataset, and the receiving-end AC side disturbance response dataset. In the disturbance response dataset of the sending-end AC side, extract the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending-end converter station; In the DC-side disturbance response dataset, extract the transient overshoot peak data and the time constant data of fluctuation decay of the DC voltage at the receiving-end converter station; In the AC-side disturbance response dataset at the receiving end, the transient drop duration data of the wind turbine grid connection point voltage and the transient fluctuation energy data of the wind farm output power are extracted.

3. The method according to claim 1, characterized in that, The establishment of collaborative optimization criteria in step S3 specifically includes: Based on the transient drop envelope data and recovery trajectory oscillation mode data of the AC bus voltage of the sending-end converter station, a first performance index group is constructed. The first performance index group is used to characterize the voltage support strength and recovery stability of the sending-end converter station after disturbance. Based on the transient overshoot peak data and fluctuation decay time constant data of the DC voltage at the receiving-end converter station, a second performance index group is constructed. The second performance index group is used to characterize the voltage robustness and dynamic response speed of the DC grid when facing power surges. Based on the transient drop duration data of the grid connection point voltage of the wind turbine and the transient fluctuation energy data of the output power of the wind farm, a third performance index group is constructed. The third performance index group is used to characterize the wind farm's ability to operate without disconnecting from the grid and the stability of its power output during grid disturbances. The first, second, and third performance index groups are jointly established as the collaborative optimization criteria.

4. The method according to claim 1, characterized in that, Step S4 employs a multi-objective evolutionary optimization mechanism for iterative search, specifically including: Initialize a search population containing multiple sets of control parameters for the sending-end converter station and control parameters for the wind turbine converter, with each set of parameters as an independent search entity; Each individual in the search population is assigned to an electromagnetic transient simulation model in turn, and multiple preset disturbance conditions are simulated to obtain the actual performance values ​​of the first performance index group, the second performance index group and the third performance index group corresponding to each individual. Based on the actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy, and the sparseness of the distribution of individuals in the target space is evaluated by the crowding distance within the same level. Based on the Pareto front level and the crowding distance, select some individuals from the current population for crossover and mutation operations to generate a new generation of search population, and repeat the above process until the preset iteration termination condition is met.

5. The method according to claim 1, characterized in that, Step S5 involves identifying a set of parameter combinations that achieve the optimal balance among multiple performance indicators, specifically including: From the final generation of search population that meets the iteration termination condition, select all search individuals located at the first Pareto front level to form a candidate parameter solution set; In the candidate parameter solution set, the parameter individual with the minimum degree of conflict among the first performance index group, the second performance index group and the third performance index group is identified. The minimum degree of conflict is measured by the comprehensive deviation between the actual performance value and the ideal performance value of the specific index within each performance index group. The control parameters of the sending-end converter station and the control parameters of the wind turbine converter corresponding to the identified individual parameters are determined as the final collaborative optimization result.

6. A system for collaborative optimization of control parameters of a converter station and a wind turbine converter in an offshore wind power flexible DC system for implementing the method described in any one of claims 1-5, characterized in that, include: The model building unit builds an electromagnetic transient simulation model that includes the sending-end converter station, receiving-end converter station, DC transmission line and wind farm equivalent model, based on the actual topology and component characteristics of the offshore wind power flexible DC system. The data recording unit connects to the electromagnetic transient simulation model of the model building unit. Based on various preset disturbance conditions, it combines and changes the control parameters of the sending-end converter station and the wind turbine converter in the model, and records the dynamic response process data of the key electrical quantities of the sending-end converter station and the wind turbine converter under different parameter combinations. The criteria construction unit receives dynamic response process data recorded by the data recording unit and establishes a set of collaborative optimization criteria that includes multiple performance indicators, all of which are directly mapped from the dynamic response process data. The iterative search unit receives the collaborative optimization criteria from the criterion construction unit, adopts a multi-objective evolutionary optimization mechanism, takes the control parameters of the sending-end converter station and the control parameters of the wind turbine converter as variables to be optimized, uses multiple performance indicators in the collaborative optimization criteria as optimization guides, and repeatedly calls the electromagnetic transient simulation model of the model building unit to perform iterative optimization. The result output unit receives the optimization results from the iterative search unit, and identifies a set of control parameters for the sending-end converter station and the wind turbine converter that achieves the optimal balance of multiple performance indicators from the final parameter combination. This set is then output as a collaborative optimization configuration scheme.

7. The system according to claim 6, characterized in that, When the data recording unit records dynamic response process data, it divides the dynamic response process data into sending-end AC side disturbance response dataset, DC side disturbance response dataset, and receiving-end AC side disturbance response dataset according to the location and type of disturbance. In the disturbance response dataset of the sending-end AC side, extract the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending-end converter station; In the DC-side disturbance response dataset, extract the transient overshoot peak data and the time constant data of fluctuation decay of the DC voltage at the receiving-end converter station; In the AC-side disturbance response dataset at the receiving end, the transient drop duration data of the wind turbine grid connection point voltage and the transient fluctuation energy data of the wind farm output power are extracted.

8. The system according to claim 7, characterized in that, When establishing the collaborative optimization criteria, the criterion construction unit constructs a first set of performance indicators to characterize the voltage support strength and recovery stability of the sending-end converter station based on the transient drop envelope data and the oscillation mode data of the recovery trajectory of the AC bus voltage of the sending-end converter station. Based on the transient overshoot peak data and fluctuation decay time constant data of the DC voltage at the receiving-end converter station, a second set of performance indicators is constructed to characterize the voltage robustness and dynamic response speed of the DC grid. Based on the transient drop duration data of the grid connection voltage of the wind turbine and the transient fluctuation energy data of the output power of the wind farm, a third set of performance indicators is constructed to characterize the wind farm's ability to operate without disconnecting from the grid and the stability of its power output. The first, second, and third performance index groups are jointly established as the collaborative optimization criteria.

9. The system according to claim 8, characterized in that, When the iterative search unit performs iterative search using a multi-objective evolutionary optimization mechanism, it initializes a search population containing multiple sets of control parameters for the sending-end converter station and control parameters for the wind turbine converter. Each search individual is sequentially assigned to an electromagnetic transient simulation model and the simulation is executed to obtain the actual performance values ​​of the first performance index group, the second performance index group and the third performance index group corresponding to each search individual. Based on actual performance values, all search individuals are divided into different Pareto front levels using a non-dominated ranking strategy, and the crowding distance of individuals is evaluated within the same level. Based on the Pareto front hierarchy and the distance between crowding levels, select individuals are crossovered and mutated to generate a new generation of search population until the preset iteration termination condition is met.

10. The system according to claim 9, characterized in that, When the result output unit identifies parameter combinations, it selects all search individuals located at the first Pareto front level from the final generation of search population that meets the iteration termination condition to form a candidate parameter solution set. In the candidate parameter solution set, the parameter individual with the minimum degree of conflict among the first performance index group, the second performance index group and the third performance index group is identified. The minimum degree of conflict is measured by the comprehensive deviation between the actual performance value and the ideal performance value of the specific index within each performance index group. The control parameters of the sending-end converter station and the control parameters of the wind turbine converter corresponding to the identified individual parameters are determined as the final collaborative optimization result.

Images