A method for reconstructing phase current of permanent magnet synchronous motor with high modulation ratio and three resistance sampling

Abstract

The application discloses a kind of permanent magnet synchronous motor high modulation ratio three resistance sampling phase current reconstruction method, S1 obtains preliminary sampling frame;S2. Generate boundary input data;S3. Initialize rule weight parameter set;S4. Get weighted candidate current interval set;S5. Cross-checking is carried out to weighted candidate current interval set and comparator array output, eliminate the inference branch that does not comply with hardware polarity and amplitude limit, obtain the candidate current interval set after boundary reduction, and with the reconstructed three-phase current vector of last pulse width modulation period fusion, generate the reconstructed three-phase current vector of current period;S6. Periodically execute steps S1 to S5, so that the reconstructed three-phase current vector is continuously kept error less than preset value under high modulation ratio driving condition, complete permanent magnet synchronous motor high modulation ratio three resistance sampling phase current reconstruction method real-time closed-loop control.The application realizes the steady-state fitting of current estimation process, avoids abnormal output of mutation point.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet synchronous motor technology, and in particular to a phase current reconstruction method for permanent magnet synchronous motors using high modulation ratio three-resistor sampling. Background Technology

[0002] With the development of high-performance servo drive systems, permanent magnet synchronous motors (PMSMs) are widely used in electric vehicles, elevators, and motor controllers due to their high power density, excellent control performance, and energy-saving advantages. In practical applications, obtaining accurate three-phase current information is crucial for achieving high-precision closed-loop control of the motor's operating status. Traditional current sampling methods mostly use Hall current sensors or shunt resistors to sample phase current signals and obtain values ​​through analog-to-digital converters for the controller. However, under high modulation ratio drive conditions, i.e., when the pulse width modulation signal is close to full duty cycle, the sampling window is extremely short or some phase currents are in dead zone shielding areas, resulting in a significant decrease in the integrity and measurability of the three-phase current signal, directly affecting the accuracy of current detection.

[0003] Currently, most mainstream phase current reconstruction methods rely on mathematical modeling and spatial vector relationship derivation. Under the framework of the two-resistor method or the one-resistor method, current compensation and reconstruction are performed through the logical relationship between the spatial voltage vector and the current pulse width modulation sector number. However, it is usually assumed that the current change is linear and controllable, and the disturbance effects of dead time, voltage utilization rate and inductance characteristics on the current change boundary under high modulation ratio are ignored. This makes the inference interval too large or severely distorted under high modulation ratio, making it difficult to guarantee real-time control accuracy.

[0004] In addition, although some improved algorithms using symbolic logic or fuzzy reasoning have theoretically enhanced the adaptability and robustness of the model, most of them still do not fully consider the interference factors of sampling saturation, inconsistent boundary polarity hardware conditions, and lack a dynamic evaluation mechanism for the credibility of candidate inference intervals. Therefore, in actual control, problems such as current identification deviation accumulation, modulation instability, and even drive malfunction are prone to occur.

[0005] In summary, existing technologies have significant shortcomings in the three-resistance sampling reconstruction problem of permanent magnet synchronous motors under high modulation ratios: on the one hand, traditional reconstruction methods have incomplete modeling of boundary constraints and excessively wide reasoning ranges; on the other hand, existing logical inference methods are still insufficient in terms of dynamic adaptability and reliability evaluation mechanisms, making it difficult to meet the requirements of actual industrial control systems for high-precision and robust current identification, which seriously restricts the application effect of permanent magnet synchronous motors in high-performance scenarios. There is an urgent need to propose a reconstruction method with more logical completeness and real-time adaptability to address the complex constraints of incomplete sampling and inconsistent polarity determination. Summary of the Invention

[0006] One objective of this invention is to propose a phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistor sampling. This invention achieves steady-state fitting of the current estimation process and avoids abnormal output at abrupt change points.

[0007] A phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistor sampling according to an embodiment of the present invention includes the following steps:

[0008] S1. Synchronous sampling of the three-phase current is performed using three sampling resistors to obtain a preliminary sampling frame;

[0009] S2. Within the same pulse width modulation cycle, read the current pulse width modulation sector number, DC bus voltage utilization rate and sampling saturation flag, as well as the comparator array's determination results of the polarity and amplitude threshold of the three-phase sampling signal, and generate boundary input data;

[0010] S3. Input the initial sampling frame and boundary input data into the symbolic logic reasoning module, perform rule matching using the static boundary rule base, output the first candidate current interval set, and initialize the rule weight parameter set at the same time;

[0011] S4. In the symbolic logic reasoning module, the dynamic weight adaptive layer is called to update the set of rule weight parameters based on the residual error between the reconstruction result of the previous cycle and the current preliminary sampling frame, and the first candidate current interval set is re-inferred to obtain the weighted candidate current interval set.

[0012] S5. Cross-validate the weighted candidate current interval set with the comparator array output, eliminate inference branches that do not meet the hardware polarity and amplitude limits, obtain the candidate current interval set after boundary reduction, and fuse it with the reconstructed three-phase current vector of the previous pulse width modulation cycle to generate the reconstructed three-phase current vector of the current cycle.

[0013] S6. Periodically execute steps S1 to S5 to ensure that the reconstructed three-phase current vector maintains an error less than a preset value under high modulation ratio drive conditions, thereby completing the real-time closed-loop control of the phase current reconstruction method of high modulation ratio three-resistance sampling for permanent magnet synchronous motor.

[0014] Optionally, S1 includes the following steps:

[0015] S11. Initialize the pulse width modulation control of the three-phase inverter for the permanent magnet synchronous motor and set the modulation ratio threshold γ. th When the modulation ratio γ≥γ th When the high modulation ratio observation mode is activated, the modulation ratio γ is defined as:

[0016]

[0017] in, V represents the maximum output voltage of the three-phase inverter. dc This is the DC bus voltage;

[0018] S12. In high modulation ratio observation mode, the three low-resistance sampling resistors R are respectively... a ,R b ,R c These are connected in series in the A, B, and C phase branches of the motor, respectively, and the instantaneous voltage across each sampling resistor is measured by a high-precision analog-to-digital converter module. Perform synchronous sampling;

[0019] S13. The system calculates the three-phase instantaneous current observations by dividing the voltage across the three low-resistance sampling resistors by their corresponding resistance values ​​according to Ohm's law. The instantaneous current observations are for phase A... Phase B C phase

[0020] S14. Collect the three-phase instantaneous current observation values. This constitutes the initial sampling frame within the current pulse width modulation period. Among them, t k The timestamp represents the current sampling time.

[0021] Optionally, S2 includes the following steps:

[0022] S21. In the current pulse width modulation period T k Internally read the sector number σ where the current PWM modulation is located. k ∈{1,2,3,4,5,6} is used to indicate the region to which the inverter output voltage vector belongs in the six spatial partitions. The sector number has a direct correspondence with the current current vector position.

[0023] S22. Synchronously acquire the DC bus voltage utilization rate μ for the current cycle. k DC bus voltage utilization reflects the voltage modulation depth and dead-zone shielding degree:

[0024]

[0025] in, V represents the amplitude of the average three-phase output voltage of the inverter during the k-th modulation cycle. dc Indicates the bus voltage;

[0026] S23. Identify the sampling status of the three-phase channels during the previous sampling cycle and record the set of sampling saturation flags. in, This indicates whether the analog-to-digital converter input oversaturation occurs in phase x during the k-th period. 1 indicates oversaturation and 0 indicates normal operation. x∈{a,b,c};

[0027] S24. Synchronously extract the polarity state and amplitude threshold judgment result of the current cycle three-phase instantaneous current signal to obtain the polarity discrimination code stream. in:

[0028]

[0029] in, V represents the instantaneous current observation value of phase x within the current cycle. th+ and V th- These are positive and negative threshold levels, used to define the boundary conditions of the current signal;

[0030] S25. Number the PWM sector σ for the current cycle. k Bus voltage utilization rate μ k Sampling saturation flag set Θ k and polarity discrimination bitstream Π k Integrate to form the boundary input data set B k .

[0031] Optionally, S3 includes the following steps:

[0032] S31. Transfer the initial sampling frame F init With boundary input data set B k Input the extended symbolic logic reasoning module, which includes a hypergraph static boundary rule base R = {R1, R2, ..., R...} constructed with multidimensional boundary information. m};

[0033] S32. Based on the PWM sector number σ of the current pulse width modulation period k DC bus voltage utilization rate μ k and the stator phase inductance L of the permanent magnet synchronous motor s Within the symbolic logic reasoning module, a sector-voltage gradient parameter set is constructed. The sector-voltage gradient parameter set is used to characterize the maximum voltage change trend of each phase current under given sector and bus conditions. The magnitude of the voltage gradient of each phase current is proportional to the DC bus voltage utilization rate based on the unit voltage direction sign of the phase within the current PWM sector number, and is also affected by the inductance characteristics.

[0034]

[0035] in, Let x be the unit voltage direction coefficient of phase x within the PWM sector;

[0036] S33. Combine the sector-voltage gradient parameter set with the preset dead time Δt dzCalculate the maximum disturbance amplitude of the three-phase current within the current modulation period, constructing a current boundary change vector. This vector describes the maximum increase or decrease in the three-phase current caused by the change in driving voltage during the dead time, and is defined as the amplitude of the current change of phase A. Phase B C phase

[0037] S34. Based on three-phase instantaneous current observations and the corresponding current boundary change vector Construct the first candidate current interval set for the current period The candidate range for the phase A current is extended upwards and downwards by a maximum disturbance amplitude from the current instantaneous phase A current value. The structure is similar for phase B and phase C, forming corresponding candidate intervals. The set of candidate intervals reflects the physically feasible intervals of the three-phase current under the current dead zone effect and voltage utilization conditions, and is used to limit the boundary range of the reasoning path in the symbolic logic reasoning module.

[0038]

[0039] S35. Subset of candidate rules Initialize the set of rule weight parameters Ω k :

[0040]

[0041] Among them, Ω k This represents the subset of candidate rules selected by the symbolic logic reasoning module under the current pulse width modulation period. The generated set of rule weight parameters is initialized, and each weight parameter within the set of rule weight parameters... Corresponding rule R i The contribution of ‖Θ to the reliability of phase current inference within this period k ‖ represents the Hamming weight of the sampling saturation flag set, i.e., the number of three-phase sampling channels in the current period that are in a saturated state, α i Representation rule R i For the sensitivity factor of bus voltage utilization, β i Representation rule R i The suppression coefficient for sampling saturation.

[0042] Optionally, the method for constructing the hypergraph static boundary rule base includes:

[0043] Based on pulse width modulation geometric constraints, the PWM sector number σ is defined. k Polarity discrimination bitstream Π k Construct a sector-polarity correspondence matrix by combining the legal combinations of these elements;

[0044] Based on voltage-current physical constraints, combined with DC bus voltage utilization μ k Dead time Δt dz and motor stator phase inductance L s Establish a current gradient model to describe the possible current variation within a unit dead time.

[0045] Based on the sampling saturation flag set Θ k Define the reasonability state of each phase channel and generate a sampling availability mask;

[0046] Boundary information is mapped to a six-dimensional discrete state node graph, where edges represent constraint coupling relationships between rules and nodes represent feasible phase current variation modes. This is ultimately fused to form a hypergraph-based static boundary rule library R, ​​possessing topological connectivity and electrical interpretability. Each rule R within this hypergraph-based static boundary rule library... i Mapped to a specific candidate current range structure.

[0047] Optionally, S4 includes the following steps:

[0048] S41. In the current pulse width modulation period T k Within, call the previous cycle T. k-1 The generated three-phase current reconstruction vector And compare it with the three-phase instantaneous current observations in the initial sampling frame of the current cycle. Differential calculations are performed to obtain the three-phase residual error set, which reflects the difference between the reconstruction result of the previous cycle and the current actual sampling.

[0049] S42. A subset of candidate rules for symbolic logic inference within the current pulse width modulation period based on the three-phase residual error set. Each rule undergoes residual response analysis to generate an error response function for the current period. The error response function is used to characterize the contribution of each rule to the current residual in the reconstruction of the previous period. The larger the residual amplitude of the error response function, the lower the fitness of the rule in the previous period.

[0050] S43. An exponentially regularized decreasing mechanism is used to update the set of rule weight parameters from the previous period. During the update process, a regularization adjustment coefficient is introduced to obtain the set of rule weight parameters Ω for the current period. k The set of rule weight parameters is used to characterize the contribution of each rule to the credibility of symbolic logic reasoning in the current period;

[0051] S44. After completing the update of the rule weight parameter set, use the rule weight parameter set Ω for the current period. k For the first candidate current interval set Perform a re-inference operation, and use a weighted fusion strategy to weight the confidence of each phase current across multiple candidate intervals, outputting a set of weighted candidate current intervals.

[0052] Optionally, S5 includes the following steps:

[0053] S51. Read the current pulse width modulation period T k The set of weighted candidate current intervals within the data is obtained, and the polarity discrimination code stream provided by the comparator array in the boundary input data set is obtained simultaneously.

[0054] S52. For each phase current x∈{a,b,c}, in the weighted candidate current interval set In this paper, an interval elimination rule based on symbolic constraints is constructed, and the polarity discriminant value is used as the basis. Define the effective polarity direction and generate a set of candidate current intervals after boundary consistency verification.

[0055] S53. For each phase current x, in the set of candidate current intervals Built-in sliding weighted fitting window The fitting window incorporates the reconstructed current value from the previous cycle, with the current cycle interval as its core. As a regression reference, a weighted minimum deviation fitting strategy is employed:

[0056]

[0057] in, This represents the reconstructed current value of phase x in the current cycle, where w1 and w2 are the fitting weighting coefficients for balancing historical trends and inferred stability, and δ i This represents the deviation factor between the candidate point and the current center interval position, controlling the error growth caused by the center offset of the interval;

[0058] S54. Combine the three-phase current fitting results and output the reconstructed three-phase current vector for the current cycle.

[0059] When the candidate current interval set If the data is non-empty and the fitting error is less than a set threshold, reconstruct the three-phase current vector. This is the point with the smallest deviation within the interval;

[0060] When the candidate current interval set If the value is empty, or the fitting error exceeds the limit, then the three-phase current vector values ​​reconstructed from the previous week will be used. As compensation output.

[0061] Optionally, the effective polarity direction is defined by the following rule:

[0062] When polarity is used to determine the bitstream Retain the portion of the interval whose upper bound is greater than 0;

[0063] When polarity is used to determine the bitstream The portion of the interval whose lower bound is less than 0 is retained;

[0064] When polarity is used to determine the bitstream Keep all intervals;

[0065] The elimination operation is performed separately on each of the three phases, generating a set of candidate current intervals after boundary consistency verification.

[0066] The beneficial effects of this invention are:

[0067] (1) This invention constructs a hypergraph static boundary rule library, integrates pulse width modulation sector number, voltage utilization rate, sampling saturation state and polarity information to construct a multi-dimensional rule node graph, realizes logical modeling of complex factors such as dead zone, inductance characteristics and sampling polarity changes, and discretizes these complex physical boundary conditions into logical node constraints. On this basis, a set of rule weight parameters is introduced and dynamically updated through a periodic residual feedback mechanism, which effectively avoids the problem of misinference caused by insufficient generalization of the initial rule library, and enables the candidate current interval to have higher inference confidence within the physical boundary conditions.

[0068] (2) The present invention introduces a periodic residual error feedback mechanism. Based on the difference between the three-phase current reconstruction result of the previous cycle and the preliminary sampling frame of the current cycle, a residual response function is constructed to perform fitness analysis on the symbolic reasoning rules and dynamically adjust the set of rule weight parameters so that the reasoning process is more biased towards the rule path with smaller residual and stronger fitting ability, effectively overcoming the logic drift problem caused by polarity reversal and signal masking.

[0069] (3) Before the inference result is output, the present invention performs symbol consistency elimination on the candidate current interval based on the polarity discrimination code stream, further eliminating invalid intervals with logical conflicts. Then, it constructs a sliding fitting window by combining the reconstructed current vector of the previous cycle, and generates the current estimation result of the current cycle through the weighted minimum deviation fitting strategy. Taking into account the historical trend, the confidence of the current interval and the stability of the sampling boundary, the steady-state fitting of the current estimation process is realized, avoiding abnormal output at the sudden point. Attached Figure Description

[0070] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0071] Figure 1This is a flowchart of a phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistor sampling, as proposed in this invention. Detailed Implementation

[0072] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0073] refer to Figure 1 A phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistor sampling includes the following steps:

[0074] S1. Synchronous sampling of the three-phase current is performed using three sampling resistors to obtain a preliminary sampling frame;

[0075] S2. Within the same pulse width modulation cycle, read the current pulse width modulation sector number, DC bus voltage utilization rate and sampling saturation flag, as well as the comparator array's determination results of the polarity and amplitude threshold of the three-phase sampling signal, and generate boundary input data;

[0076] S3. Input the initial sampling frame and boundary input data into the symbolic logic reasoning module, perform rule matching using the static boundary rule base, output the first candidate current interval set, and initialize the rule weight parameter set at the same time;

[0077] S4. In the symbolic logic reasoning module, the dynamic weight adaptive layer is called to update the set of rule weight parameters based on the residual error between the reconstruction result of the previous cycle and the current preliminary sampling frame, and the first candidate current interval set is re-inferred to obtain the weighted candidate current interval set.

[0078] S5. Cross-validate the weighted candidate current interval set with the comparator array output, eliminate inference branches that do not meet the hardware polarity and amplitude limits, obtain the candidate current interval set after boundary reduction, and fuse it with the reconstructed three-phase current vector of the previous pulse width modulation cycle to generate the reconstructed three-phase current vector of the current cycle.

[0079] S6. Periodically execute steps S1 to S5 to ensure that the reconstructed three-phase current vector maintains an error less than a preset value under high modulation ratio drive conditions, thereby completing the real-time closed-loop control of the phase current reconstruction method of high modulation ratio three-resistance sampling for permanent magnet synchronous motor.

[0080] In this embodiment, S1 includes the following steps:

[0081] S11. Initialize the pulse width modulation control of the three-phase inverter for the permanent magnet synchronous motor and set the modulation ratio threshold γ. th When the modulation ratio γ≥γ th When the high modulation ratio observation mode is activated, the modulation ratio γ is defined as:

[0082]

[0083] in, V represents the maximum output voltage of the three-phase inverter. dc This is the DC bus voltage;

[0084] S12. In high modulation ratio observation mode, the three low-resistance sampling resistors R are respectively... a ,R b ,R c These are connected in series in the A, B, and C phase branches of the motor, respectively, and the instantaneous voltage across each sampling resistor is measured by a high-precision analog-to-digital converter module. Perform synchronous sampling;

[0085] S13. The system calculates the three-phase instantaneous current observations by dividing the voltage across the three low-resistance sampling resistors by their corresponding resistance values ​​according to Ohm's law. The instantaneous current observations are for phase A... Phase B C phase

[0086] S14. Collect the three-phase instantaneous current observation values. This constitutes the initial sampling frame within the current pulse width modulation period. Among them, t k The timestamp represents the current sampling time.

[0087] In this embodiment, S2 includes the following steps:

[0088] S21. In the current pulse width modulation period T k Internally read the sector number σ where the current PWM modulation is located. k ∈{1,2,3,4,5,6} is used to indicate the region to which the inverter output voltage vector belongs in the six spatial partitions. The sector number has a direct correspondence with the current current vector position.

[0089] S22. Synchronously acquire the DC bus voltage utilization rate μ for the current cycle. k DC bus voltage utilization reflects the voltage modulation depth and dead-zone shielding degree:

[0090]

[0091] in, V represents the amplitude of the average three-phase output voltage of the inverter during the k-th modulation cycle. dc Indicates the bus voltage;

[0092] S23. Identify the sampling status of the three-phase channels during the previous sampling cycle and record the set of sampling saturation flags. in, This indicates whether the analog-to-digital converter input oversaturation occurs in phase x during the k-th period. 1 indicates oversaturation and 0 indicates normal operation. x∈{a,b,c};

[0093] S24. Synchronously extract the polarity state and amplitude threshold judgment result of the current cycle three-phase instantaneous current signal to obtain the polarity discrimination code stream. in:

[0094]

[0095] in, V represents the instantaneous current observation value of phase x within the current cycle. th+ and V th- These are positive and negative threshold levels, used to define the boundary conditions of the current signal;

[0096] S25. Number the PWM sector σ for the current cycle. k Bus voltage utilization rate μ k Sampling saturation flag set Θ k and polarity discrimination bitstream Π k Integrate to form the boundary input data set B k .

[0097] In this embodiment, S3 includes the following steps:

[0098] S31. Transfer the initial sampling frame F init With boundary input data set B k Input the extended symbolic logic reasoning module, which includes a hypergraph static boundary rule base R = {R1, R2, ..., R...} constructed with multidimensional boundary information. m};

[0099] S32. Based on the PWM sector number σ of the current pulse width modulation period k DC bus voltage utilization rate μ k and the stator phase inductance L of the permanent magnet synchronous motor s Within the symbolic logic reasoning module, a sector-voltage gradient parameter set is constructed. The sector-voltage gradient parameter set is used to characterize the maximum voltage change trend of each phase current under given sector and bus conditions. The magnitude of the voltage gradient of each phase current is proportional to the DC bus voltage utilization rate based on the unit voltage direction sign of the phase within the current PWM sector number, and is also affected by the inductance characteristics.

[0100]

[0101] in, is the unit voltage direction coefficient of phase x within the PWM sector;

[0102] S33. Combine the sector-voltage gradient parameter set with the preset dead time Δt dz Calculate the maximum disturbance amplitude of the three-phase current within the current modulation period, constructing a current boundary change vector. This vector describes the maximum increase or decrease in the three-phase current caused by the change in driving voltage during the dead time, and is defined as the amplitude of the current change of phase A. Phase B C phase

[0103] S34. Based on three-phase instantaneous current observations and the corresponding current boundary change vector Construct the first candidate current interval set for the current period The candidate range for the phase A current is extended upwards and downwards by a maximum disturbance amplitude from the current instantaneous phase A current value. The structure is similar for phase B and phase C, forming corresponding candidate intervals. The set of candidate intervals reflects the physically feasible intervals of the three-phase current under the current dead zone effect and voltage utilization conditions, and is used to limit the boundary range of the reasoning path in the symbolic logic reasoning module.

[0104]

[0105] S35. Subset of candidate rules Initialize the set of rule weight parameters Ω k :

[0106]

[0107] Among them, Ω k This represents the subset of candidate rules selected by the symbolic logic reasoning module under the current pulse width modulation period. The generated set of rule weight parameters is initialized, and each weight parameter within the set of rule weight parameters... Corresponding rule R i The contribution of ‖Θ to the reliability of phase current inference within this period k ‖ represents the Hamming weight of the sampling saturation flag set, i.e., the number of three-phase sampling channels in the current period that are in a saturated state, α i Representation rule R i For the sensitivity factor of bus voltage utilization, β i Representation rule R i The suppression coefficient for sampling saturation.

[0108] In this embodiment, the method for constructing the hypergraph static boundary rule base includes:

[0109] Based on pulse width modulation geometric constraints, the PWM sector number σ is defined.k Polarity discrimination bitstream Π k Construct a sector-polarity correspondence matrix by combining the legal combinations of these elements;

[0110] Based on voltage-current physical constraints, combined with DC bus voltage utilization μ k Dead time Δt dz and motor stator phase inductance L s Establish a current gradient model to describe the possible current variation within a unit dead time.

[0111] Based on the sampling saturation flag set Θ k Define the reasonability state of each phase channel and generate a sampling availability mask;

[0112] Boundary information is mapped to a six-dimensional discrete state node graph, where edges represent constraint coupling relationships between rules and nodes represent feasible phase current variation modes. This is ultimately fused to form a hypergraph-based static boundary rule library R, ​​possessing topological connectivity and electrical interpretability. Each rule R within this hypergraph-based static boundary rule library... i Mapped to a specific candidate current range structure.

[0113] In this embodiment, S4 includes the following steps:

[0114] S41. In the current pulse width modulation period T k Within, call the previous cycle T. k-1 The generated three-phase current reconstruction vector And compare it with the three-phase instantaneous current observations in the initial sampling frame of the current cycle. Differential calculations are performed to obtain the three-phase residual error set, which reflects the difference between the reconstruction result of the previous cycle and the current actual sampling.

[0115] S42. A subset of candidate rules for symbolic logic inference within the current pulse width modulation period based on the three-phase residual error set. Each rule undergoes residual response analysis to generate an error response function for the current period. The error response function is used to characterize the contribution of each rule to the current residual in the reconstruction of the previous period. The larger the residual amplitude of the error response function, the lower the fitness of the rule in the previous period.

[0116] S43. An exponentially regularized decreasing mechanism is used to update the set of rule weight parameters from the previous period. During the update process, a regularization adjustment coefficient is introduced to obtain the set of rule weight parameters Ω for the current period. k The set of rule weight parameters is used to characterize the contribution of each rule to the credibility of symbolic logic reasoning in the current period;

[0117] S44. After completing the update of the rule weight parameter set, use the rule weight parameter set Ω for the current period. k For the first candidate current interval set Perform a re-inference operation, and use a weighted fusion strategy to weight the confidence of each phase current across multiple candidate intervals, outputting a set of weighted candidate current intervals.

[0118] In this embodiment, S5 includes the following steps:

[0119] S51. Read the current pulse width modulation period T k The set of weighted candidate current intervals within the data is obtained, and the polarity discrimination code stream provided by the comparator array in the boundary input data set is obtained simultaneously.

[0120] S52. For each phase current x∈{a,b,c}, in the weighted candidate current interval set In this paper, an interval elimination rule based on symbolic constraints is constructed, and the polarity discriminant value is used as the basis. Define the effective polarity direction and generate a set of candidate current intervals after boundary consistency verification.

[0121] S53. For each phase current x, in the set of candidate current intervals Built-in sliding weighted fitting window The fitting window incorporates the reconstructed current value from the previous cycle, with the current cycle interval as its core. As a regression reference, a weighted minimum deviation fitting strategy is employed:

[0122]

[0123] in, This represents the reconstructed current value of phase x in the current cycle, where w1 and w2 are the fitting weighting coefficients for balancing historical trends and inferred stability, and δ i This represents the deviation factor between the candidate point and the current center interval position, controlling the error growth caused by the center offset of the interval;

[0124] S54. Combine the three-phase current fitting results and output the reconstructed three-phase current vector for the current cycle.

[0125] When the candidate current interval set If the data is non-empty and the fitting error is less than a set threshold, reconstruct the three-phase current vector. This is the point with the smallest deviation within the interval;

[0126] When the candidate current interval set If the value is empty, or the fitting error exceeds the limit, then the three-phase current vector values ​​reconstructed from the previous week will be used. As compensation output.

[0127] In this embodiment, the effective polarity direction is defined as follows:

[0128] When polarity is used to determine the bitstream Retain the portion of the interval whose upper bound is greater than 0;

[0129] When polarity is used to determine the bitstream The portion of the interval whose lower bound is less than 0 is retained;

[0130] When polarity is used to determine the bitstream Keep all intervals;

[0131] The elimination operation is performed separately on each of the three phases, generating a set of candidate current intervals after boundary consistency verification.

[0132] Example 1:

[0133] Company A encountered a problem when debugging the performance of a new type of electric injection molding machine drive system: the permanent magnet synchronous motor frequently triggered overcurrent protection under high-speed and high-load conditions. The modulation ratio of this model of motor has been operating at an extremely high level for many years, remaining stable between 0.93 and 0.97, which is a typical high modulation ratio application scenario. Company engineer Liang Weiqing found that under this modulation ratio, the traditional current reconstruction method using two resistors to calculate the current frequently fails due to insufficient sampling window or polarity inference errors, leading to incorrect judgments by the control system, which in turn causes abnormal current limiting and interference feedback, seriously restricting the continuity and stability of equipment operation.

[0134] To address this issue, the project team decided to introduce this invention into its development platform and arranged a three-day system comparison test. The test location was the automation equipment debugging room on the third floor of Building A, Hengtu Intelligent Control Experimental Center, Bao'an District, Shenzhen. The TMS320F28379D was used as the main control chip, equipped with a self-developed three-channel current sampling module.

[0135] The project team completed the system initialization configuration, including motor parameter input, current sampling channel calibration, and PWM modulation settings. The three-phase sampling resistors were all 0.005 ohms, the sampling rate was 10kHz, and the test object was a permanent magnet synchronous motor with a rated current of 10 amps, a DC bus voltage of 540 volts, and a load of 0.7 times the rated torque.

[0136] Formal comparative experiments were then conducted, with modulation scores set to three levels: 0.89, 0.94, and 0.97. Under the same operating conditions, the traditional two-resistor inverse solution method and the algorithm of this invention were run respectively, and key index data were recorded. For ease of unified analysis, all data were synchronously collected to the backend server via the CAN port and recorded in CSV format.

[0137] Taking the B group experiment with a modulation ratio of 0.94 as an example, a total of 840 sets of current data samples were collected during the experiment, which lasted for about 1 hour. The traditional algorithm failed to handle the effects of certain polarity jumps and dead time boundaries, resulting in a maximum current reconstruction error of 1.78 amperes and an average error squared value of 0.523 amperes squared. Under the same conditions, the maximum error of the method of this invention was controlled within 0.88 amperes, and the average error was 0.176 amperes. The error fluctuation was only one-third of that of the traditional scheme.

[0138] At a specific moment, the system detected a large fluctuation in the C-phase current. Traditional methods, due to misjudging the polarity direction, identified the negative current as the positive direction, and the reconstructed value deviated from the actual value by more than 9 amperes, directly triggering the incorrect current limiting logic and causing the system to speed limit. However, the method of this invention correctly identifies the polarity of the current interval by combining boundary rule logic with the sampling upper and lower limit fluctuation range judgment, and outputs a current value close to the actual value, so the system operates without errors.

[0139] The project team further compared the frequency of controller action interference caused by current reconstruction misjudgment during continuous operation of the two methods. During 20 minutes of continuous operation, the traditional method had 11 malfunctions when the modulation ratio exceeded 0.95 under high load, some of which occurred during high-speed operation and caused equipment stoppage. However, after using the method of the present invention, the number of malfunctions was reduced to 2, and both were edge samples of brief voltage jitter that were not fully identified, which did not affect the normal operation of the equipment.

[0140] In addition, to further verify the universality of the algorithm under different working modes, the project team constructed three sets of AI training datasets, which came from three typical working conditions: low-speed constant torque, high-speed climbing section, and frequent load disturbance, totaling 1,850 samples. These were applied to the fitting parameter adjustment of the rule weight adaptive layer. The training results showed that after three rounds of training iterations, the sample set errors stabilized and converged to 0.084, 0.143, and 0.097, respectively, demonstrating strong adaptability and learning efficiency.

[0141] In terms of operational stability, the method of this invention exhibits significant advantages. Under extreme operating conditions with bus voltage fluctuations as high as 20%, the fluctuation of the three-phase current output curve is reduced by more than 41% compared to the traditional solution, resulting in a smoother drive system response and effectively reducing system oscillation problems caused by error accumulation.

[0142] After the experiment, the project team summarized the performance indicators as follows:

[0143] Table 1. Data comparison between the method of this invention and the traditional algorithm.

[0144]

[0145]

[0146] The experimental results fully verified that the method of the present invention has good current reconstruction accuracy, polarity identification robustness and real-time reasoning ability under high modulation ratio and non-ideal sampling conditions. The company subsequently decided to embed the algorithm into its new generation T900 series industrial motor controller and include it in the default configuration of products shipped in the first quarter of 2025.

[0147] This invention constructs a hypergraph-based static boundary rule base, fusing pulse width modulation sector numbers, voltage utilization, sampling saturation state, and polarity information to build a multi-dimensional rule node graph. This enables logical modeling of complex factors such as dead zones, inductance characteristics, and sampling polarity changes. These complex physical boundary conditions are discretized into logical node constraints. Based on this, a set of rule weight parameters is introduced and dynamically updated through a periodic residual feedback mechanism. This effectively avoids misinference problems caused by insufficient generalization of the initial rule base, enabling candidate current intervals to have higher inference confidence within the physical boundary conditions.

[0148] This invention introduces a periodic residual error feedback mechanism. Based on the difference between the three-phase current reconstruction result of the previous cycle and the preliminary sampling frame of the current cycle, a residual response function is constructed to perform fitness analysis on the symbolic inference rules. The set of rule weight parameters is dynamically adjusted so that the inference process is more biased towards the rule path with smaller residuals and stronger fitting ability, effectively overcoming the logic drift problem caused by polarity reversal and signal masking.

[0149] Before outputting the inference results, this invention performs symbol consistency elimination on the candidate current interval based on the polarity discrimination code stream, further eliminating invalid intervals with logical conflicts. Then, it constructs a sliding fitting window by combining the reconstructed current vector of the previous cycle, and generates the current estimation result of the current cycle through a weighted minimum deviation fitting strategy. By comprehensively considering historical trends, current interval confidence, and sampling boundary stability, it achieves steady-state fitting of the current estimation process and avoids abnormal output caused by abrupt changes.

[0150] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for reconstructing phase current using high modulation ratio three-resistance sampling in a permanent magnet synchronous motor, characterized in that, Includes the following steps: S1. Synchronous sampling of the three-phase current is performed using three sampling resistors to obtain a preliminary sampling frame; S2. Within the same pulse width modulation cycle, read the current pulse width modulation sector number, DC bus voltage utilization rate and sampling saturation flag, as well as the comparator array's determination results of the polarity and amplitude threshold of the three-phase sampling signal, and generate boundary input data; S3. Input the initial sampling frame and boundary input data into the symbolic logic reasoning module, perform rule matching using the static boundary rule base, output the first candidate current interval set, and initialize the rule weight parameter set at the same time; S4. In the symbolic logic reasoning module, the dynamic weight adaptive layer is called to update the set of rule weight parameters based on the residual error between the reconstruction result of the previous cycle and the current preliminary sampling frame, and the first candidate current interval set is re-inferred to obtain the weighted candidate current interval set. S5. Cross-validate the weighted candidate current interval set with the comparator array output, eliminate inference branches that do not meet the hardware polarity and amplitude limits, obtain the candidate current interval set after boundary reduction, and fuse it with the reconstructed three-phase current vector of the previous pulse width modulation cycle to generate the reconstructed three-phase current vector of the current cycle. S6. Periodically execute steps S1 to S5 to ensure that the reconstructed three-phase current vector maintains an error less than a preset value under high modulation ratio drive conditions, thereby completing the real-time closed-loop control of the phase current reconstruction method of high modulation ratio three-resistance sampling for permanent magnet synchronous motor.

2. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 1, characterized in that, S1 includes the following steps: S11. Initialize the pulse width modulation control of the three-phase inverter for the permanent magnet synchronous motor and set the modulation ratio threshold γ. th When the modulation ratio γ≥γ th At that time, the high modulation ratio observation mode is activated; S12. In high modulation ratio observation mode, the three low-resistance sampling resistors R are respectively... a ,R b ,R c These are connected in series in the A, B, and C phase branches of the motor, respectively, and the instantaneous voltage across each sampling resistor is measured by a high-precision analog-to-digital converter module. Perform synchronous sampling; S13. The system calculates the three-phase instantaneous current observations by dividing the voltage across the three low-resistance sampling resistors by their corresponding resistance values ​​according to Ohm's law. The instantaneous current observations are for phase A... Phase B C phase S14. Collect the three-phase instantaneous current observation values. This constitutes the initial sampling frame within the current pulse width modulation period. Among them, t k The timestamp represents the current sampling time.

3. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 2, characterized in that, S2 includes the following steps: S21. In the current pulse width modulation period T k Internally read the sector number σ where the current PWM modulation is located. k ∈{1,2,3,4,5,6} is used to indicate the region to which the inverter output voltage vector belongs in the six spatial partitions. The sector number has a direct correspondence with the current current vector position. S22. Synchronously acquire the DC bus voltage utilization rate μ for the current cycle. k DC bus voltage utilization reflects the voltage modulation depth and the degree of dead zone shielding. S23. Identify the sampling status of the three-phase channels during the previous sampling cycle and record the set of sampling saturation flags. in, This indicates whether the analog-to-digital converter input oversaturation occurs in phase x during the k-th period. 1 indicates oversaturation and 0 indicates normal operation. x∈{a,b,c}; S24. Synchronously extract the polarity state and amplitude threshold judgment result of the current cycle three-phase instantaneous current signal to obtain the polarity discrimination code stream. in: in, V represents the instantaneous current observation value of phase x within the current cycle. th+ and V th- These are positive and negative threshold levels, used to define the boundary conditions of the current signal; S25. Number the PWM sector σ for the current cycle. k Bus voltage utilization rate μ k Sampling saturation flag set Θ k and polarity discrimination bitstream Π k Integrate to form the boundary input data set B k .

4. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 3, characterized in that, S3 includes the following steps: S31. Transfer the initial sampling frame F init With boundary input data set B k Input the extended symbolic logic reasoning module, which includes a hypergraph static boundary rule base R = {R1, R2, ..., R...} constructed with multidimensional boundary information. m }; S32. Based on the PWM sector number σ of the current pulse width modulation period k DC bus voltage utilization rate μ k and the stator phase inductance L of the permanent magnet synchronous motor s Within the symbolic logic reasoning module, a sector-voltage gradient parameter set is constructed. The sector-voltage gradient parameter set is used to characterize the maximum voltage change trend of each phase current under given sector and bus conditions. The magnitude of the voltage gradient of each phase current is proportional to the DC bus voltage utilization rate based on the unit voltage direction sign of the phase within the current PWM sector number, and is also affected by the inductance characteristics. S33. Combine the sector-voltage gradient parameter set with the preset dead time Δt dz Calculate the maximum disturbance amplitude of the three-phase current within the current modulation period, constructing a current boundary change vector. This vector describes the maximum increase or decrease in the three-phase current caused by the change in driving voltage during the dead time, and is defined as the amplitude of the current change of phase A. Phase B C phase S34. Based on three-phase instantaneous current observations and the corresponding current boundary change vector Construct the first candidate current interval set for the current period The candidate range for the phase A current is extended upwards and downwards by a maximum disturbance amplitude from the current instantaneous phase A current value. The structure is similar to that of phase B and phase C, forming corresponding candidate intervals. The set of candidate intervals reflects the physically feasible intervals of the three-phase current under the current dead zone effect and voltage utilization conditions, and is used to limit the boundary range of the reasoning path in the symbolic logic reasoning module. S35. Subset of candidate rules Initialize the set of rule weight parameters Ω k : Among them, Ω k This represents the subset of candidate rules selected by the symbolic logic reasoning module under the current pulse width modulation period. The generated set of rule weight parameters is initialized, and each weight parameter within the set of rule weight parameters... Corresponding rule R i The contribution of ‖Θ to the reliability of phase current inference within this period k ‖ represents the Hamming weight of the sampling saturation flag set, i.e., the number of three-phase sampling channels in the current period that are in a saturated state, α i Representation rule R i For the sensitivity factor of bus voltage utilization, β i Representation rule R i The suppression coefficient for sampling saturation.

5. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 4, characterized in that, The method for constructing the hypergraph static boundary rule base includes: Based on pulse width modulation geometric constraints, the PWM sector number σ is defined. k Polarity discrimination bitstream Π k Construct a sector-polarity correspondence matrix by combining the legal combinations of these elements; Based on voltage-current physical constraints, combined with DC bus voltage utilization μ k Dead time Δt dz and motor stator phase inductance L s Establish a current gradient model to describe the possible current variation within a unit dead time. Based on the sampling saturation flag set Θ k Define the reasonability state of each phase channel and generate a sampling availability mask; Boundary information is mapped to a six-dimensional discrete state node graph, where edges represent constraint coupling relationships between rules and nodes represent feasible phase current variation modes. This is ultimately fused to form a hypergraph-based static boundary rule library R, ​​possessing topological connectivity and electrical interpretability. Each rule R within this hypergraph-based static boundary rule library... i Mapped to a specific candidate current range structure.

6. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 4, characterized in that, S4 includes the following steps: S41. In the current pulse width modulation period T k Within, call the previous cycle T. k-1 The generated three-phase current reconstruction vector And compare it with the three-phase instantaneous current observations in the initial sampling frame of the current cycle. Differential calculations are performed to obtain the three-phase residual error set, which reflects the difference between the reconstruction result of the previous cycle and the current actual sampling. S42. A subset of candidate rules for symbolic logic inference within the current pulse width modulation period based on the three-phase residual error set. Each rule undergoes residual response analysis to generate an error response function for the current period. The error response function is used to characterize the contribution of each rule to the current residual in the reconstruction of the previous period. The larger the residual amplitude of the error response function, the lower the fitness of the rule in the previous period. S43. An exponentially regularized decreasing mechanism is used to update the set of rule weight parameters from the previous period. During the update process, a regularization adjustment coefficient is introduced to obtain the set of rule weight parameters Ω for the current period. k The set of rule weight parameters is used to characterize the contribution of each rule to the credibility of symbolic logic reasoning in the current period; S44. After completing the update of the rule weight parameter set, use the rule weight parameter set Ω for the current period. k For the first candidate current interval set Perform a re-inference operation, and use a weighted fusion strategy to weight the confidence of each phase current across multiple candidate intervals, outputting a set of weighted candidate current intervals.

7. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 6, characterized in that, S5 includes the following steps: S51. Read the current pulse width modulation period T k The set of weighted candidate current intervals within the data is obtained, and the polarity discrimination code stream provided by the comparator array in the boundary input data set is obtained simultaneously. S52. For each phase current x∈{a,b,c}, in the weighted candidate current interval set In this paper, an interval elimination rule based on symbolic constraints is constructed, and the polarity discriminant value is used as the basis. Define the effective polarity direction and generate a set of candidate current intervals after boundary consistency verification. S53. For each phase current x, in the set of candidate current intervals Built-in sliding weighted fitting window The fitting window incorporates the reconstructed current value from the previous cycle, with the current cycle interval as its core. As a regression reference, a weighted minimum deviation fitting strategy is employed: in, This represents the reconstructed current value of phase x in the current cycle, where w1 and w2 are the fitting weighting coefficients for balancing historical trends and inferred stability, and δ i This represents the deviation factor between the candidate point and the current center interval position, controlling the error growth caused by the center offset of the interval; S54. Combine the three-phase current fitting results and output the reconstructed three-phase current vector for the current cycle. When the candidate current interval set If the data is non-empty and the fitting error is less than a set threshold, reconstruct the three-phase current vector. This is the point with the smallest deviation within the interval; When the candidate current interval set If the value is empty, or the fitting error exceeds the limit, then the three-phase current vector values ​​reconstructed from the previous week will be used. As compensation output.

8. The phase current reconstruction method for a permanent magnet synchronous motor with high modulation ratio three-resistance sampling according to claim 7, characterized in that, The effective polarity direction is defined by the following rule: When polarity is used to determine the bitstream Retain the portion of the interval whose upper bound is greater than 0; When polarity is used to determine the bitstream The portion of the interval whose lower bound is less than 0 is retained; When polarity is used to determine the bitstream Keep all intervals; The elimination operation is performed separately on each of the three phases, generating a set of candidate current intervals after boundary consistency verification.

Images