Electromagnetic compatibility optimization method and system for ship shaft frequency converter

Abstract

The invention relates to the technical field of interference suppression, in particular to an electromagnetic compatibility optimization method and system for a ship shaft frequency converter, and the method comprises the steps: carrying out the term-by-term judgment of sensitive interference energy, current errors and switching loss after the discretization and jump limitation based on a frequency sequence, and screening out a boundary-crossing sequence, the genetic algorithm executes cross combination and amplitude superposition in the residual sequence, a new frequency set is formed through transformation and rearrangement, frequency points are distributed in a plurality of intervals to be reorganized, the switch state is predicted and updated one by one through Kalman filtering, observation residual errors are corrected in real time, and the frequency points are reorganized. The harmonic distortion rate, the radiation intensity and the instruction change amplitude are synchronously output by taking the time index as a sequence, the deviation is compensated at the amplitude and phase level, the harmonic index and the radiation index are controlled in an agreed range, the switching loss and the instruction change are within threshold values, and a complete closed loop covering frequency generation, harmonic correction and sequence updating is formed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of interference suppression technology, and in particular to an electromagnetic compatibility optimization method and system for a ship shaft-mounted frequency converter. BACKGROUND

[0002] The field of interference suppression technology aims to achieve electromagnetic compatibility of electronic devices and systems by controlling interference sources, blocking propagation paths, and improving the anti-interference of sensitive devices. The core goal is to reduce the conducted interference current and the intensity of the radiated electromagnetic field below the standard limit, enhance the stable operation ability of the device under the action of the external electromagnetic field, and ensure the long-term reliable operation of the communication, navigation, control and power system in a complex electromagnetic environment.

[0003] The electromagnetic compatibility optimization method for a ship shaft-mounted frequency converter aims to reduce the conducted interference current and the intensity of the radiated electromagnetic field generated by the frequency converter during operation, control the harmonic current content and voltage distortion rate within the range specified by the electromagnetic compatibility standard, and reduce the interference coupling to the ship navigation, communication and control equipment, to ensure the stable operation of the ship electrical system under different working conditions, and to improve the reliability and service life of the equipment.

[0004] The existing technology uses fixed rules and static parameters in the direction of interference source control, propagation path blocking and sensitive device anti-interference improvement. Spectrum evaluation is based on average indicators and offline testing. There is a lack of fine differentiation and threshold tracking for communication frequency bands and navigation frequency bands under dynamic conditions and load fluctuations. Frequency scheduling is often carried out with fixed switching frequency or wideband spread spectrum. There is a risk of spectrum energy aggregation in some sensitive intervals. Coupling evaluation focuses on port conduction indicators, and the response to changes in cable layout and coupling coefficient is insufficient, resulting in a sudden increase in radiation intensity caused by local arrangement changes. The execution window lacks sequence replacement and recording mechanism, and parameter correction is delayed, resulting in a long stay in an unfavorable frequency interval, which makes communication reception sensitivity decline and navigation link intermittent error and control command jitter more likely to occur. Filters need to have a larger margin, resulting in additional volume and loss. Switching devices have increased thermal stress and decreased efficiency during long-term operation. SUMMARY

[0005] The purpose of the present application is to solve the problems existing in the prior art and to provide an electromagnetic compatibility optimization method and system for a ship shaft-mounted frequency converter.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: an electromagnetic compatibility optimization method for a ship shaft-mounted frequency converter, comprising the following steps: S1: Based on the electromagnetic monitoring parameter group, complete time sequence segmentation and amplitude calculation. High frequency amplitude is compared with communication frequency band, low frequency amplitude is compared with navigation frequency band, and switching loss threshold is determined together to generate interference threshold distribution results; S2: based on the interference threshold distribution result, complete motor terminal frequency sequence discrete and jump limit, calculate sensitive interference energy and switch loss and current error, remove out-of-bound sequence by threshold comparison, adopt genetic algorithm, cross combination and amplitude transformation of the remaining sequence, generate frequency optimization set; S3: based on the frequency optimization set, complete harmonic component extraction and cable coupling matching, simulate the next sampling period, adopt Kalman filter, enumerate switch state to calculate harmonic distortion rate, radiation intensity and instruction change amplitude, determine and compensate disturbance deviation according to threshold, establish preferred scheduling sequence under constraints; S4: based on the preferred scheduling sequence under constraints, complete time window division and numbering, load switch frequency in interval and record distribution, replace the previous sequence with a new sequence at the end of the window, and obtain periodic sequence execution data; S5: based on the periodic sequence execution data, extract filter inductance and motor terminal voltage amplitude and determine sensitive frequency band, shorten the window and increase the weight when the amplitude is close to the threshold, adjust the factor to update the preferred scheduling sequence under constraints, and obtain adaptive sequence update result.

[0007] As a further scheme of the application, the interference threshold distribution result includes communication interference amplitude, navigation interference amplitude and loss threshold judgment value, the frequency optimization set includes frequency distribution sequence, interference energy index and current error index, the preferred scheduling sequence under constraints includes harmonic constraint sequence, radiation constraint sequence and instruction constraint sequence, the periodic sequence execution data includes interval number, frequency execution value and replacement record, and the adaptive sequence update result includes window length adjustment value, weight adjustment value and updated scheduling sequence.

[0008] As a further scheme of the application, the specific steps for generating the interference threshold distribution result are: Based on the electromagnetic monitoring parameter group, first do time sequence segmentation, convert the current amplitude and voltage amplitude of each segment into a matrix, compare the high frequency amplitude with the communication frequency threshold value in turn and record the difference, and generate the communication interference comparison result; Based on the communication interference comparison result, compare the low frequency amplitude with the navigation frequency threshold value item by item, add the difference and compare it with the switch loss threshold value, form a judgment mark, and generate the interference threshold distribution result.

[0009] As a further scheme of the application, the specific steps for generating the frequency optimization set are: Based on the interference threshold distribution result, do segment processing on the motor terminal frequency sequence and mark the time index, adopt discretization method to split the continuous frequency into independent points, and call the amplitude parameter and sensitive frequency range comparison for each discrete point and accumulate the interference energy, generate the interference energy calculation result; Based on the interference energy calculation results, the current amplitude and voltage amplitude of the corresponding discrete points are called to accumulate the switching loss, and the loss value and current error value are stored in the judgment table at the same time. Each item in the table is compared with the set threshold one by one and the out-of-bounds sequence is eliminated to generate the threshold screening sequence result. Based on the threshold-screened sequence results, a genetic algorithm is used to call the frequency points in the remaining sequences to cross-combine in pairs and superimpose the amplitude parameters. Then, the combined frequency sequences are transformed and rearranged to output a new sequence set and archive it, generating a frequency optimization set.

[0010] As a further aspect of the present invention, the genetic algorithm first uses the remaining frequency points in the threshold screening sequence results as initial individuals, and parameterizes the frequency points according to a set encoding method. Then, within the population size range, several frequency point pairs are selected as crossover objects, and crossover operations are performed on the selected frequency points pair by pair to generate new combined frequency sequences. During the crossover process, the corresponding amplitude parameters are superimposed. Then, transformation operations are performed on the generated combined sequences, including frequency point rearrangement and amplitude redistribution, and new individuals are formed according to the index order. Finally, the transformed frequency sequences are archived and incorporated into the new generation sequence set to form a frequency optimization set.

[0011] As a further aspect of the present invention, the specific steps for generating the preferred scheduling sequence under the aforementioned constraints are as follows: Based on the frequency optimization set, the harmonic components of each sequence are extracted and a matrix relationship is established with the cable coupling parameters. The harmonic frequencies and coupling values ​​are matched item by item using each item in the matrix and the results are stored to generate harmonic coupling matching results. Based on the harmonic coupling matching results, Kalman filtering is used to enumerate each switch state and calculate the corresponding harmonic distortion rate one by one within the sampling period. The radiation intensity and the change amplitude of the command are synchronously entered into the reference table. The data in the table are sorted by time index and output uniformly to generate the sampling period calculation results. Based on the sampling period calculation results, the distortion rate and intensity values ​​are compared with the threshold entries and the cases exceeding the limits are marked. The deviation sequence is compensated and corrected at the amplitude and phase levels, and the overall sequence set is updated to establish the optimal scheduling sequence under constraints.

[0012] As a further aspect of the present invention, the Kalman filter first uses the harmonic coupling matching results within the sampling period as input observations to establish state variables, including switch state indices, harmonic frequencies, and corresponding current and voltage amplitudes. Based on the system state transition equation and the observation equation, the predicted state and the predicted covariance are calculated. Then, the observed data and the predicted state are subtracted to obtain the observation residuals, and the Kalman gain is calculated in conjunction with the covariance matrix. Next, the predicted state is corrected using the Kalman gain and updated to the current state estimate, while the covariance matrix is ​​updated. Finally, the above prediction and update steps are performed one by one for all switch states within the sampling period, the corresponding harmonic distortion rate is calculated and recorded, the radiation intensity and command change amplitude are synchronously stored in a lookup table, and the results are uniformly output after being sorted by time index to form the sampling period calculation results.

[0013] As a further aspect of the present invention, the specific steps for generating the periodic sequence execution data are as follows: Based on the preferred scheduling sequence under the constraints, time windows are divided and assigned numbers. The switching frequencies of the corresponding intervals are loaded into the sequence list one by one and a mapping relationship is established to generate the interval frequency loading result. Based on the frequency loading results of the intervals, the frequency distribution of each interval is recorded and archived. When the window expires, the previous sequence is replaced with a newly generated sequence and written into the storage unit to obtain the periodic sequence execution data.

[0014] As a further aspect of the present invention, the specific steps for generating the adaptive sequence update result are as follows: Based on the periodic sequence execution data, the amplitude of the filtered inductor current and the amplitude of the motor terminal voltage are extracted. The amplitudes of the sensitive frequency bands are compared with the thresholds one by one and summarized into a set to generate the sensitive frequency band determination results. Based on the sensitive frequency band determination results, the time window length is shortened and new parameters are set. At the same time, the weight factor of the sensitive frequency band is increased and applied to the optimal scheduling sequence under constraints. The updated sequence is then output to obtain the adaptive sequence update result.

[0015] An electromagnetic compatibility (EMC) optimization system for a marine shaft-driven frequency converter, the system being used to execute the aforementioned EMC optimization method for the marine shaft-driven frequency converter, the system comprising: Electromagnetic threshold generation module: Based on the electromagnetic monitoring parameter group, the current and voltage amplitudes of the sampling period are segmented, the high-frequency amplitude is compared with the communication frequency band threshold and the difference is recorded, the low-frequency amplitude is compared with the navigation frequency band threshold and the difference is accumulated, the accumulated value is compared with the switching loss threshold and a mark is generated to obtain the interference threshold distribution; Frequency set optimization module: Based on the interference threshold distribution, the frequency sequence of motor terminals is discretized and jumps are restricted. The interference energy, switching loss and current error are calculated by calling the current and voltage amplitudes. Out-of-bounds sequences are eliminated. For the retained frequency points, the genetic algorithm is called to cross-combine and superimpose the amplitude. The combined sequences are rearranged and archived to obtain the frequency optimization set. Harmonic scheduling establishment module: Based on the frequency optimization set, extract the harmonic components and cable coupling parameters to establish a matrix relationship, match and record the data item by item, enumerate the switch state during the sampling period, call Kalman filter to calculate the prediction and observation residuals and update the state, record the harmonic distortion rate and radiation intensity and command change amplitude according to the time index, correct the deviation at the amplitude and phase levels, and obtain the constrained scheduling sequence. Window sequence execution module: Based on the constraint scheduling sequence, it divides the time window and numbers it, loads and records the distribution of the interval switching frequency, replaces the old sequence with the new sequence and stores it when the window ends, and obtains the periodic execution data; Adaptive update module: Based on periodic execution data, extract the amplitude of the filtered inductor current and the amplitude of the motor terminal voltage, compare the amplitude of the sensitive frequency band with the threshold and form a set, shorten the window and adjust the weight for frequency bands close to the threshold, apply the factor to the constraint scheduling sequence, and obtain the adaptive sequence update.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, after the discrete and jump restrictions based on the frequency sequence, the sensitive interference energy, current error and switching loss are judged item by item and the out-of-bounds sequence is screened out. The genetic algorithm performs crossover combination and amplitude superposition in the remaining sequence, and then forms a new frequency set through transformation and rearrangement. The frequency points are reorganized in multiple intervals, the energy concentration decreases, the spectral peaks are dispersed, the out-of-bounds probability is reduced, and the scheduling sequence has high stability. In this invention, the switching states are predicted and updated one by one through Kalman filtering, the observation residuals are corrected in real time, the harmonic distortion rate and radiation intensity and the command change amplitude are all output synchronously with time index as the sequence, the deviation is compensated at the amplitude and phase levels, the control process within the cycle maintains continuity and constraint consistency, the harmonic index and radiation index are controlled within the agreed range, and the switching loss and command change are within the threshold, forming a complete closed loop of coverage frequency generation, harmonic correction and sequence update. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the workflow of the present invention; Figure 2 This is a system flowchart of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Example 1

[0019] Please see Figure 1 This invention provides a technical solution: an electromagnetic compatibility optimization method for a ship shaft-driven frequency converter, comprising the following steps: S1: Based on the electromagnetic monitoring parameter set, complete the time sequence segmentation and amplitude calculation, compare the high frequency amplitude with the communication frequency band, compare the low frequency amplitude with the navigation frequency band, and generate the interference threshold distribution result together with the switch loss threshold determination; S2: Based on the interference threshold distribution results, the frequency sequence of the motor terminal is discretized and jump limit is set. The sensitive interference energy, switching loss and current error are calculated for each sequence. The out-of-bounds sequences are eliminated by threshold comparison. The genetic algorithm is used to cross-combine and transform the amplitude of the remaining sequences to generate the frequency optimization set. S3: Based on the frequency optimization set, complete the harmonic component extraction and cable coupling matching, simulate the next sampling period, use Kalman filtering, enumerate the switch state to calculate the harmonic distortion rate and radiation intensity and command change amplitude, judge and compensate for disturbance deviation according to the threshold, and establish the optimal scheduling sequence under constraints. S4: Based on the optimal scheduling sequence under constraints, complete the time window division and numbering, load the switching frequency in the interval and record the distribution, replace the previous sequence with a new sequence when the window ends, and obtain the periodic sequence execution data; S5: Based on the periodic sequence execution data, extract the amplitude of the filter inductor and motor terminal voltage and determine the sensitive frequency band. When the amplitude is close to the threshold, shorten the window and increase the weight. Adjust the factor to update the optimal scheduling sequence under the constraints and obtain the adaptive sequence update result.

[0020] The interference threshold distribution results include communication interference amplitude, navigation interference amplitude, and loss threshold judgment value. The frequency optimization set includes frequency distribution sequence, interference energy index, and current error index. The optimal scheduling sequence under constraints includes harmonic constraint sequence, radiation constraint sequence, and instruction constraint sequence. The periodic sequence execution data includes interval number, frequency execution value, and replacement record. The adaptive sequence update results include window length adjustment value, weight adjustment value, and updated scheduling sequence.

[0021] The specific steps for generating the interference threshold distribution results are as follows: Based on the electromagnetic monitoring parameter set, time-series segmentation is first performed, and the current amplitude and voltage amplitude of each segment are converted into a matrix. The high-frequency amplitude is then compared with the communication frequency band threshold and the difference is recorded to generate communication interference comparison results. Based on the communication interference comparison results, the low-frequency band amplitude is compared with the navigation frequency band threshold item by item. The difference is accumulated and compared with the switching loss threshold to form a judgment mark and generate the interference threshold distribution result. Based on the electromagnetic monitoring parameter set, a segmented matrix generation method is used to process the current amplitude and voltage amplitude within the sampling period. The segmented operation is called to divide the sampling period into time segments of 50 points in length. The current amplitude and voltage amplitude of the segments are reconstructed into a two-dimensional matrix in a 10-row, 5-column manner. The amplitude of the first 3 rows of the high-frequency band of the matrix is ​​extracted and compared with the communication frequency band threshold point by point. The high-frequency band amplitude is subtracted from the communication frequency band threshold point by point, and the difference is recorded. The difference is stored in the vector set in sequence to generate the communication interference comparison result. Based on the communication interference comparison results, a threshold accumulation judgment method is used to compare and judge the low-frequency band amplitude. The amplitude values ​​of the 7th to 10th rows of the low-frequency band are extracted and compared with the navigation frequency band threshold point by point. The point-by-point operation is to subtract the navigation frequency band threshold from the low-frequency band amplitude and record the difference. The set of differences is accumulated to obtain the cumulative difference. The cumulative difference is compared with the switching loss threshold. The comparison rule is to mark it as 1 when it is greater than the threshold and mark it as 0 when it is less than the threshold. The marking results are stored in the judgment table. Combining the previous vector set and the contents of the judgment table, the interference threshold distribution results are generated.

[0022] The specific steps for generating the frequency optimization set are as follows: Based on the interference threshold distribution results, the frequency sequence of the motor terminals is segmented and labeled with time index. The continuous frequency is split into independent points using a discretization method. For each discrete point, the amplitude parameter is compared with the sensitive frequency band range and the interference energy is accumulated to generate the interference energy calculation result. Based on the interference energy calculation results, the current amplitude and voltage amplitude of the corresponding discrete points are called to accumulate the switching loss, and the loss value and current error value are stored in the judgment table at the same time. The items in the table are compared with the set threshold one by one and the out-of-bounds sequence is eliminated to generate the threshold screening sequence result. Based on the threshold-screened sequence results, a genetic algorithm is used to call the frequency points in the remaining sequence to cross-combine in pairs and superimpose the amplitude parameter. Then, the combined frequency sequence is transformed and rearranged to output a new sequence set and archive it, generating a frequency optimization set. Based on the interference threshold distribution results, the frequency sequence of the motor terminals is processed by the sequence discretization method. The continuous frequency sequence is segmented according to the sampling period and labeled with the time index. Each frequency value is split into independent points by interval cutting. The amplitude parameter of each independent point is called and compared with the sensitive frequency band range item by item. The comparison operation is to subtract the upper and lower limit thresholds of the sensitive frequency band from the amplitude and record the difference. All differences are accumulated to generate the cumulative energy value and output the interference energy calculation result. Based on the interference energy calculation results, a threshold comparison method is used to process the current amplitude and voltage amplitude of discrete points. The current amplitude and voltage amplitude corresponding to each discrete point are multiplied point by point and accumulated to obtain the switching loss value. The loss value and the corresponding current error value are stored in the judgment table. Each data item in the judgment table is compared with the set threshold one by one. The comparison rule is that if the data is greater than the threshold, it is marked as out of bounds, and if the data is less than or equal to the threshold, it is marked as normal. After removing all out-of-bounds sequences, the threshold filtering sequence result is output. Based on the threshold-screened sequence results, a genetic algorithm is used to combine the remaining frequency points. In the initialization of the genetic algorithm, the population size is set to 100, the number of iterations is 50, the crossover probability is 0.8, and the mutation probability is 0.05. The remaining frequency points are called from the threshold-screened sequence results, and crossover and combination operations are performed on pairs. The crossover method is single-point crossover, that is, a crossover point position is randomly generated and the corresponding segments are swapped. After the combination is completed, the amplitude parameters of the two frequency points are summed to obtain a new amplitude value. Then, a transformation operation is performed on the newly generated frequency sequence. The transformation method is to rearrange all frequency values ​​in the sequence in ascending order. Then, the transformed sequence is written into the sequence set and archived, and the frequency optimization set is output.

[0023] The genetic algorithm first uses the remaining frequency points in the threshold screening sequence as initial individuals, and then parameterizes the frequency points according to the set encoding method. Next, within the population size range, several frequency point pairs are selected as crossover objects, and crossover operations are performed on the selected frequency points one by one to generate new combined frequency sequences. During the crossover process, the corresponding amplitude parameters are superimposed. Then, transformation operations are performed on the generated combined sequences, including frequency point rearrangement and amplitude redistribution, and new individuals are formed according to the index order. Finally, the transformed frequency sequences are archived and included in the new generation sequence set to form a frequency optimization set. Genetic algorithm, according to the formula: ; in: This indicates the amplitude after optimization and adjustment. Indicates the original amplitude. Representing frequency point The combined weights, This represents the amplitude adjustment factor. Representing frequency point Error correction factor, This represents the frequency difference penalty coefficient. This indicates the difference between the combined frequency point and the target frequency point; Execution process: First, multiple frequency points are selected from the ship's shaft-driven frequency converter through cross-operation. and the corresponding amplitude The frequency point and amplitude reflect the characteristics of the frequency converter before electromagnetic compatibility optimization. Then, the combined weight of the frequency points is used. The weights measure the contribution of each frequency point to the optimization process, so as to retain key frequency points in subsequent optimizations, enhance the suppression of electromagnetic interference, and introduce an amplitude adjustment coefficient. and error correction factor The amplitude is adjusted by both factors together to make the frequency point To get closer to the target frequency, correct errors in electromagnetic compatibility performance, and ensure that the optimization results meet the predetermined electromagnetic standards, a frequency difference penalty coefficient is introduced. and differences Frequency points that deviate significantly from the target frequency are penalized to prevent excessive frequency shifts in the inverter during optimization, which could affect electromagnetic compatibility performance. Finally, through multiple generations of iteration using a genetic algorithm, the optimized set of frequency points is obtained. This reflects the optimal frequency distribution of the ship's shaft-driven frequency converter after electromagnetic compatibility optimization, resulting in stronger electromagnetic compatibility performance.

[0024] The specific steps for generating the optimal scheduling sequence under constraints are as follows: Based on the frequency optimization set, the harmonic components of each sequence are extracted and a matrix relationship is established with the cable coupling parameters. The harmonic frequency and coupling value are matched item by item in the matrix and the results are stored to generate the harmonic coupling matching result. Based on the harmonic coupling matching results, Kalman filtering is used to enumerate each switch state and calculate the corresponding harmonic distortion rate one by one within the sampling period. The radiation intensity and the change amplitude of the command are synchronously entered into the reference table. The data in the table are sorted by time index and output uniformly to generate the sampling period calculation results. Based on the sampling period calculation results, the distortion rate and intensity values ​​are compared with the threshold entries and the cases exceeding the limits are marked. Compensation and correction are performed on the deviation sequence at the amplitude and phase levels, and the overall sequence set is updated to establish the optimal scheduling sequence under constraints. Based on the frequency optimization set, a matrix construction method is used to process the harmonic components and cable coupling parameters. First, Fourier decomposition is called to extract the harmonic components of each frequency sequence and record the frequency and amplitude. Then, a matrix generation command is called to build a two-dimensional matrix with the harmonic frequency as the row index and the cable coupling parameter as the column index. The matrix entries form ordered pairs with the frequency and the corresponding coupling value. The matching is performed item by item by traversal operation. For each harmonic frequency, its amplitude parameter is called and stored as a key-value pair with the corresponding coupling value in the matrix. Finally, the harmonic coupling matching result is output. Based on the harmonic coupling matching results, Kalman filtering is used to estimate the switch state within the sampling period. The state variable vector is set to consist of the change amplitude of the harmonic distortion rate radiation intensity command. The prediction step calls the state transition matrix multiplied by the state estimate of the previous moment. The covariance prediction is obtained by multiplying the state transition matrix by the covariance matrix of the previous moment and adding it to the process noise matrix. The observation step calls the observation matrix multiplied by the predicted state to obtain the observed predicted value. The actual observed value is subtracted from the predicted observed value to form the residual. The Kalman gain is calculated by multiplying the predicted covariance matrix by the transpose of the observation matrix and dividing by the observation residual covariance. The updated state estimate is the predicted state plus the product of the Kalman gain and the residual. The updated covariance is the predicted covariance minus the product of the Kalman gain, the observation matrix, and the predicted covariance. The above steps are performed one by one for all switch states within the sampling period. The change amplitude of the harmonic distortion rate radiation intensity command is written into a lookup table and sorted by time index to generate the sampling period calculation results. Based on the sampling period calculation results, the deviation compensation method is used to correct the data. The distortion rate and radiation intensity values ​​in the sampling period calculation results are compared with the threshold table one by one. For sequences marked as exceeding the limit, the amplitude correction operation is called to perform amplitude adjustment. The original amplitude and the compensation value are added or subtracted to obtain the corrected amplitude. At the same time, the phase correction operation is called to add or subtract a fixed offset angle to the corresponding phase. After the correction is completed, the updated frequency points are rewritten into the overall sequence set to establish the optimal scheduling sequence under constraints.

[0025] Kalman filtering first uses the harmonic coupling matching results within the sampling period as input observations to establish state variables, including switch state indices, harmonic frequencies, and corresponding current and voltage amplitudes. Based on the system state transition equation and observation equation, the predicted state and prediction covariance are calculated. Then, the observed data and the predicted state are differiated to obtain the observation residuals, and the Kalman gain is calculated in combination with the covariance matrix. Next, the predicted state is corrected using the Kalman gain and updated to the current state estimate, while the covariance matrix is ​​updated. Finally, the above prediction and update steps are performed one by one for all switch states within the sampling period, and the corresponding harmonic distortion rate is calculated and recorded. The radiation intensity and command change amplitude are synchronously stored in a lookup table, sorted by time index, and output uniformly to form the sampling period calculation results. Kalman filtering, according to the formula: ; in: Indicates time Estimates of the state of the ship's shaft-driven frequency converter system. Indicates time The system state estimate, Indicates Kalman gain, Indicates time The observed values, Represents the measurement matrix. This represents the amplitude adjustment factor. Represents the observed value Error correction factor, This represents the penalty coefficient for the difference in observed values. This represents the difference between the observed value and the target value; Execution process: First, utilize The system state estimate for the current time. Predicting the electromagnetic compatibility status by comparing current observations With the predicted state Calculate the difference, that is This represents the error between the electromagnetic radiation intensity or other relevant electromagnetic data and the predicted state. Next, the Kalman gain... The weights of the current observations are adjusted based on the magnitude of the error, making the system state updates more accurate. This is achieved by increasing the amplitude adjustment coefficient. and error correction factor The accuracy of optimization is further improved by correcting the error of the observed values, and the amplitude adjustment coefficient is used. Scaling the observations to make the adjusted observations more accurate, correction factor. This reflects the error between the current observed value and the target value, further improving the accuracy of the Kalman filter by introducing a penalty coefficient. Difference between observed values Excessive observation errors are penalized, forcing the system to pay more attention to observation data that are close to the target value, avoiding distortion of the system state due to large deviations, and dynamically adjusting the system state at each moment to ensure that the ship shaft frequency converter maintains higher accuracy and stability during electromagnetic compatibility optimization.

[0026] The specific steps for generating periodic sequence execution data are as follows: Based on the optimal scheduling sequence under constraints, time windows are divided and assigned numbers. The switching frequencies of the corresponding intervals are loaded into the sequence list one by one and a mapping relationship is established to generate the interval frequency loading result. Based on the interval frequency loading results, record and archive the frequency distribution of each interval. When the window expires, replace the previous sequence with a newly generated sequence and write it into the storage unit to obtain the periodic sequence execution data. Based on the optimal scheduling sequence under constraints, the time window partitioning method is used to process the running cycle. First, the interval partitioning operation is called to divide the entire running cycle into windows with a length of 100 milliseconds. Each window is assigned a unique number ID and stored in an index table. Then, the switching frequency value is read one by one in each window and written into the sequence list. The mapping establishment method is called to establish a one-to-one mapping relationship between the number ID and the switching frequency value. The mapping relationship is recorded in a dictionary structure in the form of key-value pairs. The interval frequency loading result is output. Based on the interval frequency loading results, a sequence archiving method is used to record and update window data. At the end of each window, a frequency distribution statistics operation is called to sort the recorded switching frequency values ​​in the window and generate a frequency distribution table. The distribution table is bound to the ID number and stored in an archive file. An archive write command is called to save the file to the specified storage path. At the same time, the window expiration signal is monitored. When the window expiration is detected, a sequence replacement operation is performed to overwrite the storage unit corresponding to the previous sequence with the newly generated frequency sequence. The replacement timestamp and sequence number are recorded to obtain the periodic sequence execution data.

[0027] The specific steps for generating adaptive sequence update results are as follows: Based on the periodic sequence execution data, the amplitude of the filtered inductor current and the amplitude of the motor terminal voltage are extracted. The amplitude of the sensitive frequency band is compared with the threshold item by item and summarized into a set to generate the sensitive frequency band judgment result. Based on the sensitive frequency band determination results, the time window length is shortened and new parameters are set. At the same time, the weight factor of the sensitive frequency band is increased and applied to the optimal scheduling sequence under constraints. The updated sequence is output to obtain the adaptive sequence update result. Based on the periodic sequence execution data, the amplitude of the filter inductor current and the amplitude of the motor terminal voltage are extracted using the interval comparison method. The sampling instruction is called to read the current amplitude and voltage amplitude at each time point in the periodic data. The extracted amplitude is compared with the sensitive frequency band threshold item by item. The comparison method is to subtract the threshold from the amplitude and record the difference. All differences are written into a temporary array. The array is set to generate a judgment set, and the sensitive frequency band judgment result is output. Based on the sensitive frequency band determination results, an adaptive update method is used to correct the scheduling sequence. Frequency band information with amplitudes close to the threshold is retrieved from the set. A window adjustment operation is performed, shortening the time window length from 100 milliseconds to 80 milliseconds and setting it as a new parameter. The sensitive frequency band number and its corresponding difference are used to calculate the weighting factor using a weighting function. The weighting factor is calculated by multiplying the difference by a preset weighting constant. The weighted value is written into a weight table. Subsequently, the optimal scheduling sequence under constraints is retrieved, and the weighting factors in the table are applied to the corresponding frequency points item by item. An update command is executed on the scheduling sequence, and the updated sequence is output, yielding the adaptive sequence update result.

[0028] Please seeFigure 2 An electromagnetic compatibility (EMC) optimization system for a marine shaft-driven frequency converter is disclosed. This system is used to execute the aforementioned EMC optimization method for the marine shaft-driven frequency converter. The system includes: Electromagnetic threshold generation module: Based on the electromagnetic monitoring parameter group, the current and voltage amplitudes of the sampling period are segmented, the high-frequency amplitude is compared with the communication frequency band threshold and the difference is recorded, the low-frequency amplitude is compared with the navigation frequency band threshold and the difference is accumulated, the accumulated value is compared with the switching loss threshold and a mark is generated to obtain the interference threshold distribution; Frequency set optimization module: Based on the interference threshold distribution, the frequency sequence of motor terminals is discretized and jumps are restricted. The interference energy, switching loss and current error are calculated by calling the current and voltage amplitudes. Out-of-bounds sequences are eliminated. For the retained frequency points, the genetic algorithm is called to cross-combine and superimpose the amplitude. The combined sequences are rearranged and archived to obtain the frequency optimization set. Harmonic scheduling establishment module: Based on the frequency optimization set, extract the harmonic components and cable coupling parameters to establish a matrix relationship, match and record the data item by item, enumerate the switch state during the sampling period, call Kalman filter to calculate the prediction and observation residuals and update the state, record the harmonic distortion rate and radiation intensity and command change amplitude according to the time index, correct the deviation at the amplitude and phase levels, and obtain the constrained scheduling sequence. Window sequence execution module: Based on the constraint scheduling sequence, it divides the time window and numbers it, loads and records the distribution of the interval switching frequency, replaces the old sequence with the new sequence and stores it when the window ends, and obtains the periodic execution data; Adaptive update module: Based on periodic execution data, extract the amplitude of the filtered inductor current and the amplitude of the motor terminal voltage, compare the amplitude of the sensitive frequency band with the threshold and form a set, shorten the window and adjust the weight for frequency bands close to the threshold, apply the factor to the constraint scheduling sequence, and obtain the adaptive sequence update.

[0029] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An electromagnetic compatibility optimization method for a ship shaft-driven frequency converter, characterized in that, Includes the following steps: S1: Based on the electromagnetic monitoring parameter set, complete the time sequence segmentation and amplitude calculation, compare the high frequency amplitude with the communication frequency band, compare the low frequency amplitude with the navigation frequency band, and generate the interference threshold distribution result together with the switch loss threshold determination; S2: Based on the interference threshold distribution results, complete the discretization and jump limit of the motor terminal frequency sequence, calculate the sensitive interference energy, switching loss and current error for each sequence, remove out-of-bounds sequences by threshold comparison, and use a genetic algorithm to cross-combine and transform the amplitude of the remaining sequences to generate a frequency optimization set. S3: Based on the frequency optimization set, complete the harmonic component extraction and cable coupling matching, simulate the next sampling period, use Kalman filtering, enumerate the switch state to calculate the harmonic distortion rate and radiation intensity and command change amplitude, judge and compensate for the disturbance deviation according to the threshold, and establish the optimal scheduling sequence under constraints. The specific steps for generating the preferred scheduling sequence under the aforementioned constraints are as follows: Based on the frequency optimization set, the harmonic components of each sequence are extracted and a matrix relationship is established with the cable coupling parameters. The harmonic frequencies and coupling values ​​are matched item by item using each item in the matrix and the results are stored to generate harmonic coupling matching results. Based on the harmonic coupling matching results, Kalman filtering is used to enumerate each switch state and calculate the corresponding harmonic distortion rate one by one within the sampling period. The radiation intensity and the change amplitude of the command are synchronously entered into the reference table. The data in the table are sorted by time index and output uniformly to generate the sampling period calculation results. Based on the sampling period calculation results, the distortion rate and intensity values ​​are compared with the threshold entries and the cases exceeding the limits are marked. Compensation and correction are performed on the deviation sequence at the amplitude and phase levels, and the overall sequence set is updated to establish the optimal scheduling sequence under constraints. S4: Based on the optimized scheduling sequence under the constraints, complete the time window division and numbering, load the switching frequency of the interval and record the distribution, replace the previous sequence with a new sequence when the window ends, and obtain the periodic sequence execution data; S5: Based on the periodic sequence execution data, extract the amplitude of the filter inductor and motor terminal voltage and determine the sensitive frequency band. When the amplitude is close to the threshold, shorten the window and increase the weight. Adjust the factor to update the optimal scheduling sequence under the constraint to obtain the adaptive sequence update result.

2. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 1, characterized in that, The interference threshold distribution results include communication interference amplitude, navigation interference amplitude, and loss threshold judgment value. The frequency optimization set includes frequency distribution sequence, interference energy index, and current error index. The preferred scheduling sequence under constraints includes harmonic constraint sequence, radiation constraint sequence, and instruction constraint sequence. The periodic sequence execution data includes interval number, frequency execution value, and replacement record. The adaptive sequence update results include window length adjustment value, weight adjustment value, and updated scheduling sequence.

3. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 1, characterized in that, The specific steps for generating the interference threshold distribution result are as follows: Based on the electromagnetic monitoring parameter set, time-series segmentation is first performed, and the current amplitude and voltage amplitude of each segment are converted into a matrix. The high-frequency amplitude is then compared with the communication frequency band threshold and the difference is recorded to generate communication interference comparison results. Based on the communication interference comparison results, the low-frequency band amplitude is compared with the navigation frequency band threshold item by item. The differences are accumulated and compared with the switching loss threshold to form a judgment mark and generate the interference threshold distribution result.

4. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 1, characterized in that, The specific steps for generating the frequency optimization set are as follows: Based on the interference threshold distribution results, the frequency sequence of the motor terminals is segmented and labeled with time index. The continuous frequency is split into independent points using a discretization method. For each discrete point, the amplitude parameter is compared with the sensitive frequency band range and the interference energy is accumulated to generate the interference energy calculation result. Based on the interference energy calculation results, the current amplitude and voltage amplitude of the corresponding discrete points are called to accumulate the switching loss, and the loss value and current error value are stored in the judgment table at the same time. Each item in the table is compared with the set threshold one by one and the out-of-bounds sequence is eliminated to generate the threshold screening sequence result. Based on the threshold-screened sequence results, a genetic algorithm is used to call the frequency points in the remaining sequences to cross-combine in pairs and superimpose the amplitude parameters. Then, the combined frequency sequences are transformed and rearranged to output a new sequence set and archive it, generating a frequency optimization set.

5. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 4, characterized in that, The genetic algorithm first uses the remaining frequency points in the threshold screening sequence as initial individuals, and then parameterizes the frequency points according to the set encoding method. Next, within the population size range, several frequency point pairs are selected as crossover objects, and crossover operations are performed on the selected frequency points pair by pair to generate new combined frequency sequences. During the crossover process, the corresponding amplitude parameters are superimposed. Then, transformation operations are performed on the generated combined sequences, including frequency point rearrangement and amplitude redistribution, and new individuals are formed according to the index order. Finally, the transformed frequency sequences are archived and incorporated into the new generation sequence set to form a frequency optimization set.

6. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 1, characterized in that, The Kalman filter first uses the harmonic coupling matching results within the sampling period as input observations to establish state variables, including switch state indices, harmonic frequencies, and corresponding current and voltage amplitudes. Based on the system state transition equation and the observation equation, the predicted state and prediction covariance are calculated. Then, the observed data and the predicted state are differencing to obtain the observation residuals, and the Kalman gain is calculated using the covariance matrix. Next, the predicted state is corrected using the Kalman gain and updated to the current state estimate, while the covariance matrix is ​​updated. Finally, the above prediction and update steps are performed on all switch states one by one within the sampling period, and the corresponding harmonic distortion rate is calculated and recorded. The radiation intensity and command change amplitude are synchronously stored in a lookup table, sorted by time index, and output uniformly to form the sampling period calculation results.

7. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 1, characterized in that, The specific steps for generating the periodic sequence execution data are as follows: Based on the preferred scheduling sequence under the constraints, time windows are divided and assigned numbers. The switching frequencies of the corresponding intervals are loaded into the sequence list one by one and a mapping relationship is established to generate the interval frequency loading result. Based on the frequency loading results of the intervals, the frequency distribution of each interval is recorded and archived. When the window expires, the previous sequence is replaced with a newly generated sequence and written into the storage unit to obtain the periodic sequence execution data.

8. The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to claim 1, characterized in that, The specific steps for generating the adaptive sequence update result are as follows: Based on the periodic sequence execution data, the amplitude of the filtered inductor current and the amplitude of the motor terminal voltage are extracted. The amplitudes of the sensitive frequency bands are compared with the thresholds one by one and summarized into a set to generate the sensitive frequency band determination results. Based on the sensitive frequency band determination results, the time window length is shortened and new parameters are set. At the same time, the weight factor of the sensitive frequency band is increased and applied to the optimal scheduling sequence under constraints. The updated sequence is then output to obtain the adaptive sequence update result.

9. An electromagnetic compatibility optimization system for a ship shaft-driven frequency converter, characterized in that, The electromagnetic compatibility optimization method for a ship shaft-driven frequency converter according to any one of claims 1-8, wherein the system comprises: Electromagnetic threshold generation module: Based on the electromagnetic monitoring parameter group, the current and voltage amplitudes of the sampling period are segmented, the high-frequency amplitude is compared with the communication frequency band threshold and the difference is recorded, the low-frequency amplitude is compared with the navigation frequency band threshold and the difference is accumulated, the accumulated value is compared with the switching loss threshold and a mark is generated to obtain the interference threshold distribution; Frequency set optimization module: Based on the interference threshold distribution, the frequency sequence of motor terminals is discretized and jumps are restricted. The interference energy, switching loss and current error are calculated by calling the current and voltage amplitudes. Out-of-bounds sequences are eliminated. For the retained frequency points, the genetic algorithm is called to cross-combine and superimpose the amplitude. The combined sequences are rearranged and archived to obtain the frequency optimization set. Harmonic scheduling establishment module: Based on the frequency optimization set, extract the harmonic components and cable coupling parameters to establish a matrix relationship, match and record the data item by item, enumerate the switch state during the sampling period, call Kalman filter to calculate the prediction and observation residuals and update the state, record the harmonic distortion rate and radiation intensity and command change amplitude according to the time index, correct the deviation at the amplitude and phase levels, and obtain the constrained scheduling sequence. Window sequence execution module: Based on the constraint scheduling sequence, it divides the time window and numbers it, loads and records the distribution of the interval switching frequency, replaces the old sequence with the new sequence and stores it when the window ends, and obtains the periodic execution data; Adaptive update module: Based on periodic execution data, extract the amplitude of the filtered inductor current and the amplitude of the motor terminal voltage, compare the amplitude of the sensitive frequency band with the threshold and form a set, shorten the window and adjust the weight for frequency bands close to the threshold, apply the factor to the constraint scheduling sequence, and obtain the adaptive sequence update.

Images