Wind turbine generator grid forming converter voltage source control method based on virtual synchronous machine characteristics

Abstract

The application discloses a wind turbine grid-connection type converter voltage source control method based on virtual synchronous machine characteristics and relates to the technical field of converter control, and comprises the following steps: obtaining an active power sequence and a machine terminal frequency sequence of a wind turbine under virtual synchronous grid connection and constructing a first operation characteristic field; performing power fluctuation time sequence analysis on the first operation characteristic field to obtain a rotor cumulative speed deviation set and identifying a pitch angle dead zone boundary state set based on the rotor cumulative speed deviation set; constructing a rotor cumulative speed evolution graph model based on the boundary state set and adopting graph search to obtain a plurality of state evolution paths; performing speed deviation convergence evaluation on each path and screening an optimal path; generating an internal electromotive force amplitude adjustment sequence based on the optimal path and correcting a preset voltage reference value, so that the cumulative deviation of the rotor speed in the inertia simulation process of the virtual synchronous machine is inhibited, the overrun mutation of the rotor speed at the dead zone boundary is avoided, and the dynamic mismatch problem between the converter and the pitch angle is prevented.

Description

Technical Field

[0001] This invention relates to the field of converter control technology, and more specifically, to a voltage source control method for a wind turbine grid-type converter based on the characteristics of a virtual synchronous machine. Background Technology

[0002] Virtual synchronous grid control is a control method that simulates the inertia and damping characteristics of synchronous generators, enabling new energy power generation devices to have autonomous voltage and frequency support capabilities. It is widely used in new power networks with a high proportion of new energy integration. When wind turbines adopt virtual synchronous grid operation, the grid-type converter participates in grid frequency regulation by constructing virtual rotor motion characteristics and dynamically controls the internal potential and output voltage based on changes in active power, thereby improving the wind turbine's response to grid disturbances and operational stability. Simultaneously, the wind turbine itself maintains mechanical-side operational balance through rotor mechanical inertia and pitch angle adjustments.

[0003] However, during the operation of the virtual synchronous machine network, as the grid-type converter continuously simulates inertia, dynamic fluctuations in active power constantly affect the virtual rotor's motion, causing the rotor speed deviation of the wind turbine to gradually accumulate during continuous operation. Since the turbine terminal frequency typically has a preset dead zone, in the initial stage of rotor speed deviation accumulation, the terminal frequency may still remain within the dead zone, making it difficult for the control strategy to promptly identify the continuous deviation in the rotor's operating state. When the rotor speed deviation gradually accumulates to near the dead zone boundary, even minor operational disturbances can cause the speed to rapidly cross the dead zone boundary, triggering a sudden speed change. This results in a dynamic mismatch in time response and adjustment amplitude between the voltage regulation action and pitch angle adjustment process of the grid-type converter, thereby affecting the inertia support effect of the virtual synchronous machine and the overall operational stability of the wind turbine.

[0004] To address the above problems, this invention proposes a solution. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a voltage source control method for a wind turbine grid-type converter based on the characteristics of a virtual synchronous machine. By analyzing the operational evolution of the active power output sequence and the turbine terminal frequency sequence, the boundary state of the cumulative rotor speed deviation is identified in advance, and the voltage source internal potential amplitude adjustment process is optimized. This solves the problem that the cumulative deviation of the wind turbine rotor speed caused by the virtual synchronous machine during inertia simulation can lead to a sudden change in speed beyond the dead zone boundary, resulting in dynamic mismatch between the converter and the pitch angle.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A voltage source control method for a grid-connected converter for wind turbines based on the characteristics of a virtual synchronous machine includes the following steps: acquiring the active power output sequence and terminal frequency sequence of the wind turbine under virtual synchronous grid connection, and constructing a first operating characteristic field; performing power fluctuation time series analysis on the first operating characteristic field to obtain a rotor cumulative speed deviation set, and identifying the boundary state set of the pitch angle dead zone based on the rotor cumulative speed deviation set; constructing an evolution graph model of the current rotor cumulative speed based on the boundary state set, and traversing the evolution graph model using a graph search algorithm to obtain several state evolution paths; performing speed deviation convergence evaluation on each state evolution path, and selecting the optimal state evolution path based on the evaluation results; generating an internal potential amplitude adjustment sequence for the voltage source based on the optimal state evolution path, and correcting the preset voltage reference value of the grid-connected converter based on the internal potential amplitude adjustment sequence.

[0008] In a preferred embodiment, constructing the first operating characteristic field specifically involves: mapping the active power output sequence and the generator terminal frequency sequence to a phase plane coordinate system with rotor electric angular velocity as the horizontal axis and active power as the vertical axis to obtain several discrete operating points; connecting the several discrete operating points in the phase plane coordinate system to obtain several time-domain trajectory lines; obtaining the density distribution and tangent direction change rate of the several time-domain trajectory lines, and identifying several operating cluster regions based on the density distribution and tangent direction change rate to obtain the first operating characteristic field.

[0009] In a preferred embodiment, the step of performing power fluctuation time-series analysis on the first operating characteristic field to obtain the rotor cumulative speed deviation set specifically involves: extracting the power deviation between the active power and the rated power at each discrete operating point based on the first operating characteristic field; identifying positive and negative power fluctuation segments based on the power deviation, and calculating the integral value of the power deviation over time within each positive and negative power fluctuation segment; converting the integral value into the equivalent change in rotor angular velocity to obtain rotor speed deviation components; accumulating all rotor speed deviation components on the same time-domain trajectory line in chronological order to obtain the rotor cumulative speed deviation corresponding to each time-domain trajectory line, thus obtaining the rotor cumulative speed deviation set; extracting the time periods within the preset frequency dead zone range of the generator terminal frequency sequence from the rotor cumulative speed deviation set; using the discrete operating point corresponding to the maximum rate of change of rotor cumulative speed deviation within the time period as the boundary state point, and using the set of all boundary state points as the boundary state set.

[0010] In a preferred embodiment, the step of constructing the evolution graph model of the current rotor cumulative speed based on the boundary state set specifically involves: taking each boundary state point in the boundary state set as the starting point, obtaining the previous discrete running points along the positive and negative directions of the time-domain trajectory line in the phase plane coordinate system to obtain several first state sequences; obtaining the starting point and ending point of several first state sequences, and constructing the evolution graph model of the current rotor cumulative speed with the boundary state points as intermediate nodes. In a preferred embodiment, the step of traversing the evolution graph model using a graph search algorithm to obtain several state evolution paths specifically involves: obtaining the change in active power and the change in rotor electric angular velocity between any two adjacent discrete operating points in the evolution graph model; calculating the state transition dissipation value between any two adjacent discrete operating points based on the change in active power and the change in rotor electric angular velocity, and using this value as the edge weight of the connecting edge between the two discrete operating points; and searching the evolution graph model using the graph search algorithm for the path with the minimum cumulative edge weight from the initial discrete operating point to the final discrete operating point, thereby obtaining several state evolution paths.

[0011] In a preferred embodiment, the step of evaluating the speed deviation convergence for each state evolution path and selecting the optimal state evolution path based on the evaluation results specifically involves: extracting the rotor speed deviation variation curve from the initial discrete operating point to the final discrete operating point on each state evolution path; calculating the average value and variance of the rotor speed deviation variation curve, and determining the speed deviation deviation degree for each state evolution path based on the average value and variance; calculating the tangent slope of the rotor speed deviation variation curve at the final discrete operating point, and determining the deviation convergence value for each state evolution path based on the tangent slope; and selecting the state evolution paths based on the speed deviation deviation degree and deviation convergence value for each state evolution path to obtain the optimal state evolution path.

[0012] In a preferred embodiment, the step of calculating the average value and variance of the rotor speed deviation change curve, and determining the speed deviation degree of each state evolution path based on the average value and variance, specifically involves: collecting multiple sampling points at equal time intervals along the time axis of the rotor speed deviation change curve, and obtaining the rotor speed deviation value corresponding to each sampling point; calculating the arithmetic mean of the rotor speed deviation values ​​corresponding to all sampling points as the average deviation value; calculating the mean of the squares of the differences between the rotor speed deviation values ​​corresponding to all sampling points and the average deviation value as the variance of the deviation; comparing the average deviation value with a preset standard deviation value to obtain the average deviation ratio, and comparing the variance of the deviation with a preset standard deviation value to obtain the variance ratio; weighted summing the average deviation ratio and the variance ratio, and using the weighted summation result as the speed deviation degree of the state evolution path.

[0013] In a preferred embodiment, the step of generating an internal potential amplitude adjustment sequence for the voltage source based on the optimal state evolution path, and correcting the preset voltage reference value of the grid-type converter based on the internal potential amplitude adjustment sequence, specifically involves: extracting the rotor speed deviation corresponding to each discrete operating point in the optimal state evolution path, and obtaining the internal potential amplitude adjustment direction corresponding to each discrete operating point according to the preset correspondence between the rotor speed deviation and the internal potential amplitude; taking the initial discrete operating point of the optimal state evolution path as the starting point, determining the generator terminal frequency change rate at each discrete operating point along the path direction; when the generator terminal frequency change rate exceeds the preset change rate limit, taking the discrete operating point as the first adjustment point; obtaining the rotor speed deviation corresponding to the first adjustment point, and calculating the adjustment step size of the internal potential amplitude based on the rotor speed deviation; arranging the adjustment step sizes of each first adjustment point on the optimal state evolution path in the path order to obtain the internal potential amplitude adjustment sequence; and superimposing each adjustment step size in the internal potential amplitude adjustment sequence with the preset voltage reference value of the grid-type converter at the current moment to obtain the corrected voltage reference value.

[0014] In a preferred embodiment, the step of obtaining the rotor speed deviation corresponding to the first adjustment point and calculating the adjustment step size of the internal potential amplitude based on the rotor speed deviation specifically involves: obtaining the rotor speed deviation corresponding to the first adjustment point and comparing the rotor speed deviation with a preset deviation range to obtain the target deviation range of the rotor speed deviation; performing a mapping analysis between the preset deviation range and the adjustment step size, and calculating the reference adjustment step size corresponding to the target deviation range based on the analysis results; obtaining the adjustment step size of the time-domain trajectory line where the first adjustment point is located at the previous first adjustment point as the preceding adjustment step size; calculating the difference between the reference adjustment step size and the preceding adjustment step size; if the difference is greater than a preset step size change rate limit, then adding the step size change rate limit to the preceding adjustment step size as the adjustment step size of the first adjustment point; if the difference is less than or equal to the step size change rate limit, then using the reference adjustment step size as the adjustment step size of the first adjustment point.

[0015] The technical effects and advantages of the voltage source control method for wind turbine grid converters based on the characteristics of virtual synchronous machines in this invention are as follows: This invention acquires the active power output sequence and terminal frequency sequence of a wind turbine under a virtual synchronous grid and constructs a first operating characteristic field, enabling the operating state to be expressed in a time-series manner in the phase plane. By performing power fluctuation time-series analysis on the first operating characteristic field, it extracts the rotor cumulative speed deviation set and identifies the boundary state set of the pitch angle dead zone, allowing the cumulative rotor speed deviation caused by the dynamic effect of active power during inertia simulation of the virtual synchronous machine to be characterized in advance. Furthermore, based on the boundary state set, it constructs an evolution graph model of the current rotor cumulative speed and employs graph search. The algorithm traverses and filters the evolution path to obtain the optimal state evolution path, thereby avoiding the problem that traditional control methods cannot detect continuous speed deviation in advance at the dead zone boundary. Then, by evaluating the speed deviation convergence of the optimal state evolution path and generating an internal potential amplitude adjustment sequence, the preset voltage reference value of the grid-type converter is finally corrected, so that the voltage regulation process can follow the evolution trend of the rotor's cumulative deviation and make feedforward corrections. This effectively suppresses the speed from exceeding the limit at the dead zone boundary and avoids the dynamic mismatch problem between converter voltage regulation and pitch angle regulation. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the voltage source control method for a wind turbine grid converter based on the characteristics of a virtual synchronous machine, according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0018] Example 1, Figure 1 The present invention provides a voltage source control method for a wind turbine grid-type converter based on the characteristics of a virtual synchronous machine, comprising the following steps: S1, obtain the active power output sequence and terminal frequency sequence of the wind turbine under virtual synchronous network, and construct the first operating characteristic field; In this embodiment, the first running feature field is constructed as follows: The active power output sequence and the terminal frequency sequence are mapped to a phase plane coordinate system with rotor electric angular velocity as the horizontal axis and active power as the vertical axis to obtain several discrete operating points. Connecting several discrete running points in the phase plane coordinate system yields several time-domain trajectory lines; The density distribution and tangent direction change rate of several time-domain trajectory lines are obtained, and several running cluster regions are identified based on the density distribution and tangent direction change rate to obtain the first running feature field.

[0019] It should be noted that the active power output sequence of wind turbines under virtual synchronous grid construction refers to the time series formed by continuously recording the active power output data of wind turbines in chronological order during operation in virtual synchronous machine control mode. It reflects the dynamic changes in active power of wind turbines under wind speed variations, load disturbances, and inertia support, and can characterize the output characteristics of wind turbines participating in grid power regulation. The turbine terminal frequency sequence refers to the time series formed by the continuous change of the frequency corresponding to the turbine terminal output voltage over time. It reflects the dynamic changes in the turbine terminal frequency relative to the rated operating state during virtual synchronous control, and embodies the frequency change patterns of wind turbines during inertia response, damping adjustment, and power balance. The active power output sequence and the turbine terminal frequency sequence correspond to each other, both arranged in the same chronological order. The active power at each moment corresponds to the turbine terminal frequency at the same moment, thus jointly characterizing the dynamic operating characteristics of wind turbines under virtual synchronous grid construction, providing basic data for subsequent operating state analysis.

[0020] Furthermore, the rotor electric angular velocity corresponding to each moment is determined based on the turbine terminal frequency sequence, and its time sequence is kept consistent with the active power output sequence, so that each sampling moment corresponds to a unique set of rotor electric angular velocity and active power data. Subsequently, a phase plane coordinate system is established with rotor electric angular velocity as the horizontal axis and active power as the vertical axis. The rotor electric angular velocity corresponding to each moment is used as the horizontal axis value and active power as the vertical axis value according to the time sequence, and the system is calibrated point by point in the phase plane coordinate system. Each set of corresponding data forms an operating state point. As the data of each sampling moment is mapped sequentially, multiple operating state points distributed according to time sequence are formed. Since each operating state point corresponds to the actual operating state of the wind turbine at different sampling moments, all operating state points together constitute several discrete operating points, which are used to characterize the dynamic correspondence between active power and rotor electric angular velocity of the wind turbine during virtual synchronous grid operation.

[0021] Furthermore, according to the sampling time sequence corresponding to the discrete operating points, the discrete operating points corresponding to adjacent time points are connected sequentially, so that the operating state extends continuously according to the actual evolution process. When two adjacent discrete operating points correspond to consecutive sampling time points, they are directly connected by line segments to represent the evolution process of the operating state from the previous time point to the next time point. When there are situations such as shutdown, restart, operating mode switching, or sampling interruption during operation that cause changes in time continuity, the current connection is terminated, and the discrete operating point after time continuity is restored is used as the new starting point to continue the connection, thereby forming a new operating trajectory. Finally, each trajectory formed by continuous connection corresponds to the state evolution path of the wind turbine in one continuous operation process. All continuous operation processes form several time-domain trajectory lines to reflect the evolution process of the operating state of the wind turbine with time changes in different operating stages.

[0022] Finally, the active power output sequence and the turbine terminal frequency sequence are mapped in the same time order, and the rotor electric angular velocity and active power after the turbine terminal frequency conversion are mapped to the phase plane coordinate system as the abscissa and ordinate, respectively, to form multiple discrete operating points. Adjacent discrete operating points are connected sequentially according to the time sequence corresponding to each discrete operating point to obtain several time-domain trajectory lines reflecting the continuous change process of the operating state. The distribution of each time-domain trajectory line in different regions of the phase plane coordinate system is statistically analyzed to determine the number of times the trajectory passes through each region and the degree of clustering of operating points, thus obtaining the trajectory density distribution. The changes in tangent direction at different positions on each time-domain trajectory line are further analyzed to determine the degree of trajectory turning change. Combined with the trajectory density distribution, regions with relatively concentrated operating states in the phase plane are identified, and regions with high trajectory clustering and consistent tangent direction changes are identified as operating cluster regions. Finally, each operating cluster region and each time-domain trajectory line together form the first operating characteristic field used to describe the spatial distribution and evolution law of the wind turbine operating state.

[0023] S2, perform power fluctuation time series analysis on the first operating characteristic field to obtain the rotor cumulative speed deviation set, and identify the boundary state set of the pitch angle dead zone based on the rotor cumulative speed deviation set; In this embodiment, a power fluctuation time series analysis is performed on the first operating characteristic field to obtain the rotor cumulative speed deviation set, specifically: Based on the first operating feature field, the power deviation between the active power and the rated power at each discrete operating point is extracted; Based on the power deviation, positive and negative power fluctuation segments are identified, and the integral values ​​of the power deviation over time within the positive and negative power fluctuation segments are calculated respectively. The integral value is converted into the equivalent change in rotor angular velocity to obtain the rotor speed deviation component; The rotor speed deviation components on the same time domain trajectory line are accumulated in time order to obtain the rotor cumulative speed deviation corresponding to each time domain trajectory line, thus obtaining the rotor cumulative speed deviation set. Extract the time period when the machine end frequency sequence is within the preset frequency dead zone from the concentrated rotor cumulative speed deviation; The discrete operating point corresponding to the maximum rate of change of the rotor's cumulative speed deviation within the time period is taken as the boundary state point, and the set of all boundary state points is taken as the boundary state set.

[0024] It should be noted that each discrete operating point in the first operating characteristic field corresponds to the actual operating state of the wind turbine at a certain sampling moment, including operating information such as active power and rotor electric angular velocity at that moment. Using the rated power of the wind turbine under rated operating conditions as a unified reference, the actual active power corresponding to each discrete operating point is compared point by point to determine the degree of deviation of the actual output power from the rated power. When the actual active power is higher than the rated power, it indicates that the current operating state is in a power-increasing state; when the actual active power is lower than the rated power, it indicates that the current operating state is in a power-decreasing state. Each discrete operating point corresponds to a power deviation result. The power deviation results corresponding to all discrete operating points form a power deviation quantity, which is used to reflect the change of the active power output of the wind turbine relative to the rated operating state throughout the entire operation process, providing a basis for subsequent power fluctuation time-series analysis.

[0025] Secondly, based on the time sequence corresponding to each discrete operating point, a continuous analysis of the power deviation is performed. When the power deviation corresponding to multiple consecutive discrete operating points is above the rated power, the continuous time interval is defined as a positive power fluctuation segment; when the power deviation corresponding to multiple consecutive discrete operating points is below the rated power, the corresponding time interval is defined as a negative power fluctuation segment. If the direction of power deviation changes during operation, the position of change is used as the boundary between adjacent fluctuation segments, thus dividing the entire operation process into several consecutive positive and negative power fluctuation segments. Subsequently, the cumulative change of power deviation over time within each fluctuation segment is statistically analyzed, obtaining the integral value of power deviation over time within the corresponding fluctuation segment. This integral value reflects the cumulative effect of the continuous action of different power fluctuation stages, providing a unified characterization of instantaneous power changes and continuous changes.

[0026] Furthermore, the integral values ​​corresponding to each fluctuation segment reflect the cumulative impact of the continuous deviation of active power from the rated operating state. To facilitate the analysis of the impact of power fluctuations on the rotor operating state, the integral values ​​are converted into equivalent changes in rotor angular velocity according to the correspondence between active power changes and rotor motion states during the operation of the virtual synchronous machine. This allows the power change results to be uniformly converted to describe the rotor operating state. After the conversion, each power fluctuation segment corresponds to a rotor speed deviation component that reflects the degree of change in rotor operating state. This deviation component can characterize the cumulative impact of power changes on rotor speed within the current fluctuation segment, providing a unified data basis for subsequent analysis of the speed change trend throughout the entire operation process.

[0027] Furthermore, each time-domain trajectory line corresponds to one continuous operation process of the wind turbine. Each fluctuation segment is distributed along its corresponding time-domain trajectory line and arranged sequentially according to its actual occurrence time. Along the same time-domain trajectory line, the rotor speed deviation components corresponding to each power fluctuation segment are accumulated sequentially. This ensures that subsequent fluctuation segments retain the influence of previous fluctuation segments during the calculation process, thus obtaining the cumulative change in rotor speed deviation as the operating state evolves. Each time-domain trajectory line corresponds to a cumulative rotor speed deviation. Different time-domain trajectories generate corresponding cumulative results, and finally, the cumulative results corresponding to all time-domain trajectories are combined to form a set of cumulative rotor speed deviations, used to comprehensively reflect the cumulative evolution of rotor speed deviation under different operating stages of the wind turbine.

[0028] Furthermore, the cumulative rotor speed deviation set and the turbine terminal frequency sequence are mapped in the same chronological order. The turbine terminal frequency at each moment is judged point by point. When the turbine terminal frequency is within a pre-set frequency dead zone, it is considered that the current frequency change has not yet reached the condition for triggering a significant adjustment action, but the rotor operating state may have already undergone continuous changes. Therefore, the time interval continuously within the preset frequency dead zone is determined as the target time period. If the turbine terminal frequency exceeds the preset frequency dead zone, it is taken as the end position of the target time period or the start position of the next target time period. By extracting all continuous time intervals that meet the conditions, multiple operating segments within the frequency dead zone are obtained, in order to further analyze the boundary characteristics of the rotor operating state continuously evolving while the frequency change is not significant.

[0029] Finally, for each target time period within the preset frequency dead zone, the change process of the rotor cumulative speed deviation is analyzed along the time direction to determine the position where the rotor cumulative speed deviation changes the fastest over time, and the corresponding discrete operating point is obtained. Since this discrete operating point represents the key position where the rotor operating state gradually changes from relatively stable to significant change, it is determined as the boundary state point. All target time periods are analyzed in the same way, with each target time period corresponding to a boundary state point. Finally, all boundary state points are summarized to form a boundary state set. The boundary state set is used to characterize the key operating position where the rotor operating state of the wind turbine has changed significantly before the frequency change has significantly exceeded the dead zone range, providing key state nodes for the subsequent construction of the rotor cumulative speed evolution diagram model.

[0030] S3. Based on the boundary state set, construct the evolution graph model of the current rotor cumulative speed, and use the graph search algorithm to traverse the evolution graph model to obtain several state evolution paths. In this embodiment, an evolution graph model of the current rotor cumulative speed is constructed based on the boundary state set, specifically as follows: Starting from each boundary state point in the boundary state set, the preceding discrete running points are obtained in the phase plane coordinate system along the positive and negative directions of the time domain trajectory line, resulting in several first state sequences. Obtain the starting and ending points of several first state sequences, and construct an evolution graph model of the current rotor cumulative speed using boundary state points as intermediate nodes.

[0031] It should be noted that each boundary state point in the boundary state set corresponds to a representative state change position during the operation of the wind turbine. Therefore, each boundary state point is used as the starting position for state tracking, and state extension is performed along the time-domain trajectory line of that boundary state point in the phase plane coordinate system. Specifically, discrete operating points corresponding to each sampling time after the boundary state point are sequentially acquired along the positive direction of the time-domain trajectory line to characterize the subsequent development process of the operating state; discrete operating points corresponding to each sampling time before the boundary state point are sequentially acquired along the negative direction of the time-domain trajectory line to characterize the evolution process of the operating state before the formation of the boundary state. The original temporal relationship between the discrete operating points is maintained during the acquisition process, so that the boundary state point, together with the discrete operating points acquired in the positive and negative directions, forms a continuous state change sequence. All boundary state points in the boundary state set are processed in the same way, ultimately yielding several first state sequences, each of which fully reflects the continuous evolution of the operating state before and after the corresponding boundary state point.

[0032] Finally, the starting and ending discrete running points in the time direction of each first state sequence are determined, and the boundary state points in the corresponding first state sequences are retained as key nodes connecting the preceding and following running states. During the construction process, the starting point of each first state sequence is used as the state entry node, the ending point as the state exit node, and the boundary state points in between as state transition nodes. Connections between nodes are established according to the original time connections of the discrete running points in the first state sequence, allowing the running state to evolve continuously along the time direction. Since different first state sequences correspond to different running changes near different boundary state points, all first state sequences are incorporated into a unified state association structure. This allows the state evolution processes corresponding to multiple boundary state points to form mutually independent yet uniformly representable state associations, ultimately constructing an evolution graph model of the current rotor cumulative speed. This evolution graph model can completely describe the continuous evolution of the rotor cumulative speed from historical running states to subsequent running states near the boundary states, and provides state nodes and their connection relationships for subsequent state evolution path searching.

[0033] In this embodiment, a graph search algorithm is used to traverse the evolution graph model to obtain several state evolution paths, specifically: Obtain the change in active power and the change in rotor electric angular velocity between any two adjacent discrete operating points in the evolution diagram model; The state transition dissipation value between any two adjacent discrete operating points is calculated based on the change in active power and the change in rotor electric angular velocity, and is used as the edge weight of the connecting edge between the two discrete operating points. Based on the graph search algorithm, the path with the minimum cumulative edge weight from the initial discrete running point to the final discrete running point is searched in the evolutionary graph model, and several state evolution paths are obtained.

[0034] It should be noted that after the evolution graph model is constructed, the original connections between the discrete operating points are maintained. Each pair of directly connected discrete operating points corresponds to the operating state of the wind turbine at two adjacent sampling times. Active power data corresponding to adjacent discrete operating points are extracted sequentially along each connection edge in the evolution graph model, and the changes in active power between two consecutive discrete operating points are compared to obtain the corresponding active power change. Simultaneously, rotor electric angular velocity data corresponding to adjacent discrete operating points are extracted, and the changes in rotor electric angular velocity between two adjacent discrete operating points are analyzed to obtain the corresponding rotor electric angular velocity change. All adjacent discrete operating points in the evolution graph model are analyzed in the same way, so that each connection edge corresponds to a set of active power change and rotor electric angular velocity change, used to characterize the degree of change in operating state between adjacent sampling times, providing basic data for subsequent evaluation of the state transition process.

[0035] Furthermore, the changes in active power and rotor electric angular velocity corresponding to each connecting edge are obtained separately, and a correlation analysis is performed on the two to comprehensively reflect the degree of state change during the transition from one discrete operating point to the next. When the change in active power between adjacent discrete operating points is small and the change in rotor electric angular velocity is relatively gentle, it indicates that the operating state transition is relatively stable, and the corresponding state transition dissipation value is small. When the change in active power is large or the change in rotor electric angular velocity is rapid, it indicates that a large state adjustment occurs during the operating state transition, and the corresponding state transition dissipation value increases accordingly. After calculating the state transition dissipation value corresponding to each connecting edge, the corresponding result is assigned to the corresponding connecting edge as an edge weight, so that each connecting edge in the evolution graph model has edge weight information that can reflect the difficulty of state transition.

[0036] Finally, using the starting discrete running point corresponding to each first state sequence in the evolution graph model as the path start point and the corresponding ending discrete running point as the path end point, different running paths in the evolution graph model are searched one by one while keeping the connection relationships of each connecting edge unchanged. During the search process, the edge weights corresponding to all connecting edges traversed by each candidate path are continuously accumulated, and the magnitude of the accumulated edge weights corresponding to different candidate paths is compared. When the accumulated edge weight corresponding to a certain path is small, it indicates that the overall state change is smoother during the gradual evolution of the running state from the starting discrete running point to the ending discrete running point, and the degree of state transition dissipation is relatively low. Therefore, this path is retained as a state evolution path that meets the requirements. The same method is used to search between each starting point and ending point in the evolution graph model, so that corresponding state evolution paths are formed near different boundary states. Finally, several state evolution paths are obtained to reflect the possible evolution process of the rotor cumulative speed under different running states.

[0037] S4. Perform a speed deviation convergence evaluation on each state evolution path and select the optimal state evolution path based on the evaluation results. In this embodiment, a speed deviation convergence evaluation is performed on each state evolution path, and the optimal state evolution path is selected based on the evaluation results, specifically as follows: Extract the rotor speed deviation change curve from the initial discrete operating point to the final discrete operating point on each state evolution path; Calculate the average and variance of the rotor speed deviation change curve, and determine the speed deviation degree of each state evolution path based on the average and variance; Calculate the slope of the tangent line of the rotor speed deviation change curve at the termination discrete operating point, and determine the deviation convergence value of each state evolution path based on the tangent line slope; The optimal state evolution path is obtained by filtering the rotational speed deviation and deviation convergence value of each state evolution path.

[0038] It should be noted that after the state evolution path is determined, each path consists of multiple discrete operating points connected consecutively in chronological order, and each discrete operating point corresponds to a rotor speed deviation value. Following the connection order of the initial discrete operating point to the final discrete operating point along the state evolution path, the rotor speed deviation corresponding to each discrete operating point is extracted sequentially, maintaining the temporal correspondence between each rotor speed deviation and its corresponding sampling time. The continuously changing rotor speed deviations are correlated according to the path order, forming a rotor speed deviation variation curve reflecting the continuous change of rotor speed deviation as the operating state evolves. Each state evolution path corresponds to a rotor speed deviation variation curve, used to fully describe the changing trend of rotor speed deviation throughout the entire state evolution process, providing a basis for subsequent analysis of the operational stability of different state evolution paths.

[0039] For each state evolution path, the rotor speed deviation variation curve is analyzed as a whole, and the rotor speed deviation values ​​at each discrete operating point on the curve are statistically analyzed to obtain the average variation level of the curve throughout the entire state evolution process. Furthermore, the dispersion of each rotor speed deviation value around the average variation level is analyzed to obtain the corresponding variance. The average value reflects the degree of deviation of the overall state evolution path from the target operating state, while the variance reflects the fluctuation of the rotor speed deviation throughout the evolution process. Combining the average value and variance, the overall variation characteristics of the rotor speed deviation corresponding to the state evolution path are analyzed. When both the average value and variance are small, it indicates that the overall deviation of the path is small and the operation is relatively stable; when either the average value or variance is large, it indicates that the path has a large cumulative deviation or significant fluctuations. Based on the above analysis results, the speed deviation degree corresponding to each state evolution path is determined, giving different state evolution paths a unified deviation evaluation index and providing a comparative basis for subsequent path selection.

[0040] In the rotor speed deviation variation curves corresponding to each state evolution path, the variation trend of rotor speed deviation near the termination discrete operating point is extracted. The direction of tangent change at this position is determined based on the changes between adjacent operating states before and after the termination discrete operating point, reflecting the rate of change of rotor speed deviation at the end of the state evolution. When the rotor speed deviation change at the termination discrete operating point tends to level off, it indicates that the rotor speed deviation corresponding to the current state evolution path has gradually stabilized, and its corresponding deviation convergence degree is high. When a large variation trend is still maintained at the termination discrete operating point, it indicates that the rotor speed deviation is still in a state of continuous change, and its corresponding deviation convergence degree is low. The deviation convergence value corresponding to each state evolution path is determined based on the variation trend at the termination discrete operating point, so that the deviation convergence value can characterize the degree to which the rotor speed deviation eventually stabilizes, providing a basis for evaluating the convergence performance of different state evolution paths.

[0041] Finally, the rotational speed deviation and deviation convergence value corresponding to each state evolution path are comprehensively evaluated to achieve a unified assessment. The rotational speed deviation reflects the overall degree of deviation of the rotor speed deviation from the target operating state throughout the entire state evolution process, while the deviation convergence value reflects the ability of the rotor speed deviation to stabilize at the end of the state evolution. During path selection, state evolution paths with smaller overall deviation and better final convergence performance are prioritized, while those with larger deviation or poor convergence performance are eliminated. After a comprehensive comparison of all state evolution paths, the state evolution path that combines smaller rotational speed deviation and higher deviation convergence capability is determined as the optimal state evolution path. This ensures that the determined operating state evolution process maintains good operational stability while allowing the rotor speed deviation to stabilize quickly, thus providing a basis for the target state evolution in the subsequent generation of the internal potential amplitude adjustment sequence.

[0042] In this embodiment, the average value and variance of the rotor speed deviation change curve are calculated, and the speed deviation degree of each state evolution path is determined based on the average value and variance, specifically as follows: Along the time axis of the rotor speed deviation change curve, multiple sampling points are collected at equal time intervals, and the rotor speed deviation value corresponding to each sampling point is obtained. Calculate the arithmetic mean of the rotor speed deviation values ​​corresponding to all sampling points, and use it as the average deviation value; The mean of the squares of the differences between the rotor speed deviation values ​​corresponding to all sampling points and the average deviation value is calculated as the deviation variance. The average deviation is compared with the preset standard deviation to obtain the average deviation ratio, and the variance is compared with the preset standard deviation to obtain the variance ratio. The average deviation ratio and the variance ratio are weighted and summed, and the weighted sum is used as the speed deviation of the state evolution path.

[0043] It should be noted that, for the rotor speed deviation change curve corresponding to each state evolution path, multiple sampling positions are sequentially selected along the time axis at uniform time intervals. This ensures that the sampling points are evenly distributed throughout the entire state evolution process, covering the entire change process from the initial discrete operating point to the final discrete operating point. Each sampling point corresponds to a specific position on the change curve and also corresponds to a rotor speed deviation value at a specific operating moment. The rotor speed deviation values ​​corresponding to each sampling point are obtained sequentially in chronological order, maintaining the original time correspondence between the sampling points. This ensures that the acquired data reflects the overall change of rotor speed deviation throughout the entire state evolution process while avoiding the impact of overly concentrated or sparse local sampling on subsequent statistical analysis results. This provides a unified data foundation for subsequent calculations of the average and variance.

[0044] After extracting the rotor speed deviation values ​​corresponding to each sampling point, a comprehensive statistical analysis is performed on the rotor speed deviation values ​​corresponding to all sampling points. This ensures that the rotor speed deviations at each sampling point are included in the calculation, resulting in an average result that reflects the overall deviation level of the entire state evolution path. Since the average deviation comprehensively considers the rotor speed deviations at each sampling moment throughout the entire state evolution process, it can reflect the overall deviation degree of the state evolution path during the entire operation. A smaller average deviation indicates that the overall rotor speed deviation corresponding to the entire state evolution path remains at a low level; a larger average deviation indicates a significant overall speed deviation during operation. Finally, the overall average result obtained through statistics is used as the average deviation, providing a basic parameter for subsequent deviation evaluation.

[0045] After obtaining the average deviation, the rotor speed deviation values ​​corresponding to each sampling point are compared one by one using the average deviation as an overall reference. The degree of deviation of each sampling point relative to the overall average level is analyzed, and the deviation results corresponding to each sampling point are statistically analyzed to eliminate the influence of the mutual cancellation of positive and negative deviations on the statistical results. After completing the statistical analysis of the deviation results corresponding to all sampling points, a statistical result reflecting the dispersion of the distribution of each sampling point around the average deviation is obtained, and this statistical result is used as the deviation variance. The deviation variance can reflect the fluctuation of rotor speed deviation throughout the entire state evolution process. When the deviation variance is small, it indicates that the rotor speed deviation corresponding to each sampling point is relatively concentrated, and the change of operating state is relatively stable; when the deviation variance is large, it indicates that the rotor speed deviation fluctuates more significantly, and the operating state has a large range of changes.

[0046] To facilitate a unified evaluation of the operating characteristics corresponding to different state evolution paths, a standard deviation value for evaluating rotor speed deviation levels and a standard variance value for evaluating fluctuation levels are pre-defined and used as reference benchmarks for their respective evaluation items. The average deviation value for each state evolution path is compared with the standard deviation value to determine the overall deviation of the current state evolution path relative to the reference level, and the average deviation ratio is obtained accordingly. Simultaneously, the deviation variance is compared with the standard variance value to determine the fluctuation level of the current state evolution path relative to the reference level, and the variance ratio is obtained accordingly. Through these comparisons, both the average deviation value and the deviation variance are converted into relative results with a unified evaluation standard, thereby facilitating comprehensive comparisons between different state evolution paths and improving the consistency of subsequent evaluation results.

[0047] Based on the influence of the mean deviation ratio and variance ratio on the operating characteristics of the state evolution path, corresponding weights are assigned respectively. The mean deviation ratio primarily reflects the overall deviation level of the state evolution path, while the variance ratio primarily reflects the fluctuation degree during the operation of the state evolution path. Subsequently, the mean deviation ratio and variance ratio are comprehensively calculated according to their respective weights, allowing both overall deviation and fluctuation to participate in the evaluation of the state evolution path. After calculation, the comprehensive result is used as the speed deviation degree of the corresponding state evolution path. Since this speed deviation degree includes information on both the overall deviation degree and the fluctuation degree, it can more comprehensively reflect the overall change characteristics of the rotor speed deviation corresponding to the state evolution path, providing a unified and comparable evaluation index for subsequent optimization of the state evolution path.

[0048] S5 generates an internal potential amplitude adjustment sequence for the voltage source based on the optimal state evolution path, and corrects the preset voltage reference value of the grid-type converter based on the internal potential amplitude adjustment sequence.

[0049] In this embodiment, an internal potential amplitude adjustment sequence for the voltage source is generated based on the optimal state evolution path, and the preset voltage reference value of the grid-type converter is corrected based on the internal potential amplitude adjustment sequence, specifically as follows: Extract the rotor speed deviation corresponding to each discrete operating point in the optimal state evolution path, and obtain the internal potential amplitude adjustment direction corresponding to each discrete operating point based on the preset correspondence between the rotor speed deviation and the internal potential amplitude. Starting from the initial discrete running point of the optimal state evolution path, the rate of change of the terminal frequency at each discrete running point is determined point by point along the path direction; When the rate of change of the terminal frequency exceeds the preset rate of change limit, the discrete operating point will be used as the first adjustment point. Obtain the rotor speed deviation corresponding to the first adjustment point, and calculate the adjustment step size of the internal potential amplitude based on the rotor speed deviation. Arrange the adjustment step size of each first adjustment point on the optimal state evolution path according to the path order to obtain the internal potential amplitude adjustment sequence. The corrected voltage reference value is obtained by superimposing each adjustment step in the internal potential amplitude adjustment sequence with the preset voltage reference value of the grid converter at the current moment.

[0050] It should be noted that after determining the optimal state evolution path, the rotor speed deviation corresponding to each discrete operating point is extracted sequentially according to the state evolution order of this path, so that each discrete operating point corresponds to a rotor speed deviation value that can reflect the current operating state. A correspondence between the rotor speed deviation and the internal potential amplitude adjustment method is pre-established, and this correspondence is used as the basis for adjustment to perform matching analysis on the rotor speed deviation corresponding to each discrete operating point one by one. When the internal potential amplitude corresponding to the rotor speed deviation needs to be increased, the adjustment direction corresponding to the current discrete operating point is determined to be increasing the internal potential amplitude; when the internal potential amplitude corresponding to the rotor speed deviation needs to be decreased, the adjustment direction corresponding to the current discrete operating point is determined to be decreasing the internal potential amplitude. Through the above processing, each discrete operating point on the optimal state evolution path has a corresponding internal potential amplitude adjustment direction.

[0051] Furthermore, using the initial discrete running point of the optimal state evolution path as the starting point of the analysis, the process traverses point by point towards the terminating discrete running point according to the original connection order of each discrete running point in the path. During the traversal, the terminal frequency change corresponding to each discrete running point is obtained, and combined with the terminal frequency change process corresponding to the current discrete running point and the adjacent discrete running points, the rate of change of the terminal frequency with time is analyzed, thereby obtaining the terminal frequency change rate corresponding to each discrete running point. Since the entire analysis process maintains the original time evolution order of the optimal state evolution path, it can truly reflect the dynamic change characteristics of the terminal frequency during the continuous evolution of the operating state, providing a basis for identifying positions that need key adjustments.

[0052] After analyzing the rate of change of the terminal frequency corresponding to each discrete operating point, the rate of change of the terminal frequency corresponding to each discrete operating point is compared with a preset rate of change limit. When the rate of change of the terminal frequency corresponding to a certain discrete operating point exceeds the preset rate of change limit, it indicates that the operating state at that location is changing rapidly, and there is a significant need for frequency adjustment. Therefore, this discrete operating point is determined as the first adjustment point. For discrete operating points whose rate of change of terminal frequency does not exceed the preset rate of change limit, the original operating state is maintained, and they are not considered as key adjustment targets. After judging all discrete operating points in the optimal state evolution path in the above manner, several first adjustment points that require key internal potential amplitude adjustment can be obtained, giving the subsequent adjustment process clear target positions.

[0053] For each first adjustment point, the corresponding rotor speed deviation is extracted and used as a parameter to characterize the degree of deviation of the current operating state. Combined with pre-set adjustment rules, the required internal potential amplitude adjustment degree for the current operating state is analyzed. When the rotor speed deviation corresponding to the first adjustment point is large, it indicates a high degree of deviation from the target operating state, and a larger internal potential amplitude adjustment range is determined accordingly. Conversely, when the rotor speed deviation corresponding to the first adjustment point is small, it indicates a relatively low degree of deviation, and a smaller internal potential amplitude adjustment range is determined accordingly. After analysis, the corresponding internal potential amplitude adjustment result is expressed as the adjustment step size corresponding to that first adjustment point, ensuring that each first adjustment point has a corresponding adjustment amount, providing basic data for forming a continuous adjustment process.

[0054] After determining the adjustment step size corresponding to each first adjustment point, the original state evolution order of the optimal state evolution path is maintained. The corresponding adjustment step sizes are arranged sequentially according to the order of the first adjustment points in the path, ensuring that each adjustment step size is consistent with the evolution process of the actual operating state. Since the first adjustment points all originate from the same optimal state evolution path, there is a continuous time correspondence between the adjustment step sizes. After being arranged sequentially, a set of adjustment data that reflects the continuous adjustment process of the internal potential amplitude is formed. Finally, all the adjustment step sizes arranged in the path order are combined to form the internal potential amplitude adjustment sequence, enabling the subsequent internal potential amplitude to be adjusted sequentially according to the evolution process of the operating state, thus improving the continuity and consistency of the adjustment process.

[0055] Finally, according to the arrangement order of the adjustment step sizes in the internal potential amplitude adjustment sequence, each adjustment step size is applied to the preset voltage reference value corresponding to the current operating moment of the grid-type converter, so that the preset voltage reference value is gradually corrected according to the adjustment sequence. At each first adjustment point, the preset voltage reference value corresponding to the current operating moment is used as the basis for correction, and the voltage reference value is adjusted in combination with the corresponding adjustment step size, so that the corrected voltage reference value can reflect the actual adjustment requirements of the internal potential amplitude under the current operating state. As the internal potential amplitude adjustment sequence is executed sequentially, the voltage reference value corresponding to each operating moment is updated synchronously, and finally a corrected voltage reference value that conforms to the operating law of the optimal state evolution path is formed, so that the output voltage of the grid-type converter can better match the change process of the rotor operating state, thereby improving the voltage regulation stability and operation coordination during the virtual synchronous grid operation process.

[0056] In this embodiment, the rotor speed deviation corresponding to the first adjustment point is obtained, and the adjustment step size of the internal potential amplitude is calculated based on the rotor speed deviation, specifically as follows: Obtain the rotor speed deviation corresponding to the first adjustment point, and compare the rotor speed deviation with the preset deviation range to obtain the target deviation range of the rotor speed deviation. The preset deviation range is mapped to the adjustment step size, and the benchmark adjustment step size corresponding to the target deviation range is calculated based on the analysis results. Obtain the adjustment step size of the time domain trajectory line where the first adjustment point is located at the previous first adjustment point, and use it as the preceding adjustment step size; Calculate the difference between the baseline adjustment step size and the previous adjustment step size. If the difference is greater than the preset step size change rate limit, add the previous adjustment step size to the step size change rate limit and use it as the adjustment step size of the first adjustment point. If the difference is less than or equal to the step size change rate limit, then the reference adjustment step size is used as the adjustment step size for the first adjustment point.

[0057] It should be noted that for each first adjustment point in the optimal state evolution path, its corresponding rotor speed deviation is extracted and used as the basis for representing the degree of deviation of the current operating state from the target operating state. Several consecutive deviation intervals are pre-divided according to the range of rotor speed deviation, with each interval corresponding to a different degree of operational deviation. Subsequently, the rotor speed deviation corresponding to the first adjustment point is compared sequentially with each pre-defined deviation interval to determine its deviation range, and the deviation interval corresponding to that rotor speed deviation is determined as the target deviation interval. Through this process, different first adjustment points can be assigned to different target deviation intervals according to their respective degrees of rotor speed deviation, thus providing a unified classification basis for subsequently determining the corresponding adjustment step size.

[0058] A pre-established correspondence between each deviation range and its adjustment step size is established, ensuring that each deviation range corresponds to an adjustment step size range that meets the adjustment requirements of the current operating state. Then, using the target deviation range corresponding to the first adjustment point as the matching object, the corresponding adjustment method is searched in the pre-established correspondence, and the reference adjustment step size corresponding to the current target deviation range is determined based on the correspondence. When the target deviation range corresponds to a larger rotor speed deviation range, a relatively larger reference adjustment step size is determined; when the target deviation range corresponds to a smaller rotor speed deviation range, a relatively smaller reference adjustment step size is determined. Finally, the adjustment result obtained based on the target deviation range matching is used as the reference adjustment step size corresponding to the current first adjustment point, ensuring that the adjustment amplitude remains consistent with the degree of rotor speed deviation.

[0059] Since multiple first adjustment points on the same time-domain trajectory are distributed sequentially according to the evolution order of the operating state, when determining the adjustment step size of the current first adjustment point, the adjustment step size corresponding to the most recent first adjustment point before the current first adjustment point is extracted, and this adjustment step size is used as the preceding adjustment step size of the current first adjustment point. If the current first adjustment point is the first adjustment point to appear on the time-domain trajectory, then the previously determined initial adjustment value is used as the preceding adjustment step size. By obtaining the preceding adjustment step size, the current adjustment process can inherit the adjustment state of the previous stage, avoiding independent adjustment between adjustment points, thereby ensuring the continuity of the internal potential amplitude adjustment process and providing a reference for subsequently limiting abrupt changes in the adjustment step size.

[0060] After obtaining the reference adjustment step size and the previous adjustment step size corresponding to the current first adjustment point, the magnitude of change between the two is compared to determine whether the current adjustment requirement has changed significantly compared to the previous adjustment. When the difference between the reference adjustment step size and the previous adjustment step size exceeds the preset step size change rate limit, it indicates that directly using the reference adjustment step size for adjustment will cause a large jump in the adjustment magnitude between two adjacent first adjustment points, which is not conducive to maintaining a smooth adjustment of the internal potential amplitude. Therefore, in this case, the reference adjustment step size is not used directly. Instead, the allowable change amount of the step size change rate limit is added to the previous adjustment step size, and the result is used as the adjustment step size corresponding to the current first adjustment point. This allows the current adjustment step size to change gradually within the preset allowable range, thereby ensuring a smoother adjustment process.

[0061] Finally, when the difference between the reference adjustment step size and the previous adjustment step size does not exceed the preset step size change rate limit, it indicates that the change between the current adjustment demand and the previous adjustment state is small. Directly using the reference adjustment step size will not cause significant abrupt changes in the adjustment process. Therefore, there is no need to further limit or correct the adjustment step size. Instead, the reference adjustment step size determined based on the target deviation range is directly used as the adjustment step size corresponding to the current first adjustment point. The final adjustment step size corresponding to each first adjustment point is determined sequentially in the above manner, so that each adjustment point can not only meet the adjustment demand corresponding to the current operating state, but also maintain the continuous change between adjacent adjustment processes. This makes the internal potential amplitude adjustment sequence smoother, improves the stability of the preset voltage reference value correction process of the grid-type converter, and avoids large fluctuations in the voltage reference value due to abrupt changes in the adjustment step size.

[0062] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0063] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0064] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0065] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0066] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A voltage source control method for a wind turbine grid-type converter based on the characteristics of a virtual synchronous machine, characterized in that, Includes the following steps: Obtain the active power output sequence and terminal frequency sequence of the wind turbine under virtual synchronous grid construction, and construct the first operating characteristic field; A power fluctuation time series analysis is performed on the first operating characteristic field to obtain the rotor cumulative speed deviation set, and the boundary state set of the pitch angle dead zone is identified based on the rotor cumulative speed deviation set. An evolution graph model of the current rotor cumulative speed is constructed based on the boundary state set, and a graph search algorithm is used to traverse the evolution graph model to obtain several state evolution paths. For each state evolution path, a speed deviation convergence evaluation is performed, and the optimal state evolution path is selected based on the evaluation results. The internal potential amplitude adjustment sequence of the voltage source is generated based on the optimal state evolution path, and the preset voltage reference value of the grid converter is corrected based on the internal potential amplitude adjustment sequence.

2. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 1, characterized in that, The construction of the first operational feature field specifically involves: The active power output sequence and the terminal frequency sequence are mapped to a phase plane coordinate system with rotor electric angular velocity as the horizontal axis and active power as the vertical axis to obtain several discrete operating points. Connecting several discrete running points in the phase plane coordinate system yields several time-domain trajectory lines; The density distribution and tangent direction change rate of several time-domain trajectory lines are obtained, and several running cluster regions are identified based on the density distribution and tangent direction change rate to obtain the first running feature field.

3. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 2, characterized in that, The power fluctuation time-series analysis of the first operating characteristic field yields the rotor cumulative speed deviation set, specifically as follows: Based on the first operating feature field, the power deviation between the active power and the rated power at each discrete operating point is extracted; Based on the power deviation, positive and negative power fluctuation segments are identified, and the integral values ​​of the power deviation over time within the positive and negative power fluctuation segments are calculated respectively. The integral value is converted into the equivalent change in rotor angular velocity to obtain the rotor speed deviation component; The rotor speed deviation components on the same time-domain trajectory line are accumulated in time order to obtain the rotor cumulative speed deviation corresponding to each time-domain trajectory line, thus obtaining the rotor cumulative speed deviation set. Extract the time period when the machine end frequency sequence is within the preset frequency dead zone from the concentrated rotor cumulative speed deviation; The discrete operating point corresponding to the maximum rate of change of the rotor's cumulative speed deviation within the time period is taken as the boundary state point, and the set of all boundary state points is taken as the boundary state set.

4. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 3, characterized in that, The evolution graph model for constructing the current rotor cumulative speed based on the boundary state set is as follows: Starting from each boundary state point in the boundary state set, the preceding discrete running points are obtained in the phase plane coordinate system along the positive and negative directions of the time domain trajectory line, resulting in several first state sequences. Obtain the starting and ending points of several first state sequences, and construct an evolution graph model of the current rotor cumulative speed using boundary state points as intermediate nodes.

5. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 4, characterized in that, The graph search algorithm is used to traverse the evolutionary graph model to obtain several state evolution paths, specifically: Obtain the change in active power and the change in rotor electric angular velocity between any two adjacent discrete operating points in the evolution diagram model; The state transition dissipation value between any two adjacent discrete operating points is calculated based on the change in active power and the change in rotor electric angular velocity, and is used as the edge weight of the connecting edge between the two discrete operating points. Based on the graph search algorithm, the path with the minimum cumulative edge weight from the initial discrete running point to the final discrete running point is searched in the evolutionary graph model, and several state evolution paths are obtained.

6. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 5, characterized in that, The process of evaluating the speed deviation convergence for each state evolution path and selecting the optimal state evolution path based on the evaluation results is as follows: Extract the rotor speed deviation change curve from the initial discrete operating point to the final discrete operating point on each state evolution path; Calculate the average and variance of the rotor speed deviation change curve, and determine the speed deviation degree of each state evolution path based on the average and variance; Calculate the slope of the tangent line of the rotor speed deviation change curve at the termination discrete operating point, and determine the deviation convergence value of each state evolution path based on the tangent line slope; The optimal state evolution path is obtained by filtering the rotational speed deviation and deviation convergence value of each state evolution path.

7. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 6, characterized in that, The calculation of the average value and variance of the rotor speed deviation change curve, and the determination of the speed deviation degree for each state evolution path based on the average value and variance, specifically involves: Along the time axis of the rotor speed deviation change curve, multiple sampling points are collected at equal time intervals, and the rotor speed deviation value corresponding to each sampling point is obtained. Calculate the arithmetic mean of the rotor speed deviation values ​​corresponding to all sampling points, and use it as the average deviation value; The mean of the squares of the differences between the rotor speed deviation values ​​corresponding to all sampling points and the average deviation value is calculated as the deviation variance. The average deviation is compared with the preset standard deviation to obtain the average deviation ratio, and the variance is compared with the preset standard deviation to obtain the variance ratio. The average deviation ratio and the variance ratio are weighted and summed, and the weighted sum is used as the speed deviation of the state evolution path.

8. The voltage source control method for wind turbine grid-type converters based on virtual synchronous machine characteristics according to claim 7, characterized in that, The process of generating an internal potential amplitude adjustment sequence for the voltage source based on the optimal state evolution path, and then correcting the preset voltage reference value of the grid-type converter based on the internal potential amplitude adjustment sequence, specifically involves: Extract the rotor speed deviation corresponding to each discrete operating point in the optimal state evolution path, and obtain the internal potential amplitude adjustment direction corresponding to each discrete operating point based on the preset correspondence between the rotor speed deviation and the internal potential amplitude. Starting from the initial discrete running point of the optimal state evolution path, the rate of change of the terminal frequency at each discrete running point is determined point by point along the path direction; When the rate of change of the terminal frequency exceeds the preset rate of change limit, the discrete operating point will be used as the first adjustment point. Obtain the rotor speed deviation corresponding to the first adjustment point, and calculate the adjustment step size of the internal potential amplitude based on the rotor speed deviation. Arrange the adjustment step size of each first adjustment point on the optimal state evolution path according to the path order to obtain the internal potential amplitude adjustment sequence. The corrected voltage reference value is obtained by superimposing each adjustment step in the internal potential amplitude adjustment sequence with the preset voltage reference value of the grid converter at the current moment.

9. The voltage source control method for wind turbine grid converters based on virtual synchronous machine characteristics according to claim 8, characterized in that, The specific steps for obtaining the rotor speed deviation corresponding to the first adjustment point and calculating the adjustment step size of the internal potential amplitude based on the rotor speed deviation are as follows: Obtain the rotor speed deviation corresponding to the first adjustment point, and compare the rotor speed deviation with the preset deviation range to obtain the target deviation range of the rotor speed deviation. The preset deviation range is mapped to the adjustment step size, and the benchmark adjustment step size corresponding to the target deviation range is calculated based on the analysis results. Obtain the adjustment step size of the time domain trajectory line where the first adjustment point is located at the previous first adjustment point, and use it as the preceding adjustment step size; Calculate the difference between the baseline adjustment step size and the previous adjustment step size. If the difference is greater than the preset step size change rate limit, add the previous adjustment step size to the step size change rate limit and use it as the adjustment step size of the first adjustment point. If the difference is less than or equal to the step size change rate limit, then the reference adjustment step size is used as the adjustment step size for the first adjustment point.

Images