Adaptive control system of high-speed motor based on working condition data driving

Abstract

The application relates to the technical field of motor control, in particular to a high-speed motor adaptive control system based on working condition data driving, which is used for acquiring the stator winding end temperature and electrical excitation parameters of a high-speed motor synchronously sampled at multiple time points under a starting working condition; determining the motor total loss at each time point according to the electrical excitation parameters; determining the current heat dissipation capacity of the stator winding end and the current heat dissipation capacity variation trend according to the motor total loss and the stator winding end temperature; determining a predicted temperature peak value according to the current heat dissipation capacity, the current heat dissipation capacity variation trend and a preset maximum loss target value; determining a maximum allowable loss value according to the comparison result of the predicted temperature peak value and an end allowable temperature threshold value, and generating adaptive control parameters according to the maximum allowable loss value. The system can improve the motor power output capacity and operation reliability under the premise of ensuring the thermal safety of the stator winding end.

Description

Technical Field

[0001] This application relates to the field of motor control technology, specifically to an adaptive control system for a high-speed motor driven by operating condition data. Background Technology

[0002] Against the backdrop of the development of aerospace and high-end power equipment towards high power density, high efficiency, high reliability and high integration, the integrated starter-generator high-speed motor system has gradually become a key component in engine starting, airborne power supply and compact energy conversion units. This type of system usually consists of a starter generator, an integrated starter-generator controller and an engine mechanical connection mechanism. During the start-up phase, the controller inverts DC power into AC power to drive the starter generator to rotate the engine to offline speed. During this process, the stator winding current amplitude is large and the synchronous frequency rises rapidly, the fundamental copper loss and iron loss increase sharply, while the cooling oil pump has not yet established its rated working pressure, and the convective heat transfer is at its lowest level in the entire operation process. Under the effect of this trade-off, the stator winding end, as the area with the worst heat dissipation conditions, concentrated heat capacity and most sensitive local temperature rise, experiences a rapid temperature rise, which places extremely high demands on thermal safety monitoring.

[0003] In existing technologies, when dealing with such startup conditions, temperature data is usually collected by reading temperature measurement points embedded inside the motor, and the thermal state of the motor is judged in combination with empirical thresholds. When the empirical threshold is reached, protective actions such as power limiting or shutdown are performed.

[0004] However, due to the uneven temperature distribution and rapid thermal fluctuations at the stator winding ends, the limited number of temperature measurement points makes it difficult to fully capture the true thermal state at the ends. When the local temperature rise at the stator winding ends has already increased rapidly, the temperature data output by the limited temperature measurement points actually lags behind the actual thermal state. Setting the empirical threshold too high will cause the protection action to be delayed, putting the winding insulation at risk of overheating. Setting the empirical threshold too low may trigger power limiting too early, unnecessarily reducing the motor's torque output capability, resulting in decreased acceleration performance or insufficient load-carrying capacity. This leads to a discrepancy between thermal protection judgment and actual control requirements, making it difficult to balance the motor's power output capability and operational reliability while ensuring the thermal safety of the stator winding ends. Summary of the Invention

[0005] To address the technical problem in related technologies where the discrepancy between thermal protection judgment and actual control requirements makes it difficult to balance the power output capacity and operational reliability of the motor while ensuring the thermal safety of the stator winding ends, this application aims to provide a high-speed motor adaptive control system driven by operating condition data. The specific technical solution adopted is as follows: The high-speed motor adaptive control system based on operating condition data provided in this application includes a data acquisition module, a loss determination module, a heat dissipation analysis module, a temperature prediction module, and an adaptive control module. The data acquisition module acquires stator winding end temperatures and electrical excitation parameters of the high-speed motor simultaneously sampled at multiple moments during startup. The loss determination module determines the total motor loss at each moment based on the electrical excitation parameters. The heat dissipation analysis module determines the current heat dissipation capacity and its changing trend at the stator winding end based on the total motor loss and stator winding end temperatures at multiple moments. The temperature prediction module determines a predicted peak temperature based on the current heat dissipation capacity, its changing trend, and a preset maximum loss target value. The adaptive control module determines the maximum allowable loss value based on a comparison between the predicted peak temperature and an allowable end temperature threshold, and generates adaptive control parameters based on this maximum allowable loss value.

[0006] Optionally, the electrical excitation parameters include the instantaneous value of multiphase current, modulation ratio, and synchronization frequency. The aforementioned loss determination module is specifically used to: determine the fundamental copper loss at each moment based on the synchronization frequency, the instantaneous value of multiphase current at each moment, and the preset frequency AC resistance curve; determine the iron loss at each moment based on the synchronization frequency, the modulation ratio at each moment, and the preset frequency iron loss characteristic surface; determine the additional end loss at each moment based on the fundamental copper loss at each moment and the preset frequency correction coefficient, wherein the preset frequency correction coefficient is used to characterize the proportion of additional end loss that varies with frequency; and determine the total motor loss at each moment based on the fundamental copper loss, the iron loss at each moment, and the additional end loss at each moment.

[0007] Optionally, the loss determination module described above is further specifically used to: determine the effective value of each phase current based on the instantaneous value of the multiphase current at each moment; interpolate the single-phase AC resistance value at each moment from the preset frequency AC resistance curve based on the synchronization frequency at each moment; and determine the fundamental copper loss at each moment based on the effective value of each phase current and the single-phase AC resistance value at each moment.

[0008] Optionally, the loss determination module is further specifically used to: determine the modulation ratio at each moment as the equivalent voltage utilization rate at each moment; and determine the iron loss at each moment based on the synchronization frequency at each moment, the equivalent voltage utilization rate at each moment, and the preset frequency iron loss characteristic surface.

[0009] Optionally, the aforementioned heat dissipation analysis module is specifically used for: using the total motor loss as the independent variable and the stator winding end temperature as the dependent variable, performing linear fitting on the total motor loss and stator winding end temperature within the sliding time window at each moment to obtain the fitted straight line corresponding to each moment; using the slope of the fitted straight line at each moment as the heat dissipation capacity at each moment, and determining the heat dissipation capacity at the current moment as the current heat dissipation capacity; determining the average total motor loss within the sliding time window at each moment; and determining the trend of change of the current heat dissipation capacity based on the heat dissipation capacity at each moment and the average total motor loss at each moment.

[0010] Optionally, the aforementioned heat dissipation analysis module is further specifically used to: determine candidate values ​​for the heat dissipation capacity change trend at each moment based on the difference between the heat dissipation capacity at each moment and the heat dissipation capacity at the corresponding previous moment, and the difference between the average total motor loss at each moment and the average total motor loss at the corresponding previous moment; when a preset number of consecutive candidate values ​​for the heat dissipation capacity change trend meet a preset consistency condition, the candidate value for the heat dissipation capacity change trend at the current moment is determined as the current heat dissipation capacity change trend.

[0011] Optionally, the temperature prediction module is specifically used to: determine the loss change range within the sliding time window to which the current moment belongs based on the total motor loss at the start time and the total motor loss at the end time within the sliding time window to which the current moment belongs; determine the adaptive loss step size based on the loss change range and the number of moments contained in the sliding time window; use the total motor loss value at the current moment as the initial loss value for iteration, the stator winding end temperature at the current moment as the initial temperature value for iteration, the current heat dissipation capacity as the initial heat dissipation capacity value for iteration, and the current heat dissipation capacity change trend as the iteration trend value, and perform step iterations based on the adaptive loss step size to obtain the predicted loss value and predicted temperature value for each iteration step; when the predicted loss value is greater than or equal to the preset maximum loss target value for the first time, the predicted temperature value of this iteration step is determined as the predicted temperature peak value.

[0012] Optionally, the temperature prediction module described above is further configured to: in each iteration, determine the predicted loss value of the current iteration step by summing the predicted loss value of the previous iteration step with the adaptive loss step size; determine the heat dissipation capacity value of the current iteration step based on the adaptive loss step size, the iteration trend value, and the heat dissipation capacity value of the previous iteration step; determine the temperature increment of the current iteration step by multiplying the adaptive loss step size with the heat dissipation capacity value of the current iteration step; and determine the predicted temperature value of the current iteration step by summing the predicted temperature value of the previous iteration step with the temperature increment of the current iteration step.

[0013] Optionally, the temperature prediction module described above is further configured to: determine the reliability coefficient of the current iteration step based on the current iteration step number; determine the heat dissipation capability correction amount of the current iteration step by multiplying the adaptive loss step size, the iteration trend value, and the reliability coefficient of the current iteration step; and determine the heat dissipation capability value of the current iteration step by summing the heat dissipation capability value of the previous iteration step and the heat dissipation capability correction amount of the current iteration step.

[0014] Optionally, the aforementioned adaptive control module is specifically configured to: when the predicted peak temperature is less than or equal to the allowable end temperature threshold, determine the preset maximum loss target value as the maximum allowable loss value; when the predicted peak temperature is greater than the allowable end temperature threshold, perform a binary search with the current total motor loss value as the lower bound of the search interval and the preset maximum loss target value as the upper bound of the search interval; in each binary search, use the loss value at the midpoint of the search interval as a candidate loss value, and perform step-by-step iterative deduction based on the candidate loss value, the current heat dissipation capacity, and the current heat dissipation capacity change trend to obtain the candidate predicted temperature value; update the search interval based on the comparison result between the candidate predicted temperature value and the allowable end temperature threshold until the search interval width is less than the preset convergence tolerance; and determine the loss value at the midpoint of the converged search interval as the maximum allowable loss value.

[0015] This application has the following beneficial effects: By acquiring stator winding end temperatures and electrical excitation parameters simultaneously sampled at multiple moments during startup, a data foundation with strictly aligned timestamps is provided for subsequent thermal analysis. The input quantities for thermal drive are established by converting the electrical excitation parameters at each moment into the total motor loss at each moment. The real-time heat dissipation status is accurately reflected by extracting the current heat dissipation capacity and its changing trend. This allows for accurate prediction of the peak temperature based on the current heat dissipation capacity, its changing trend, and the preset maximum loss target value. Finally, the maximum allowable loss value is determined by comparing the predicted peak temperature with the allowable end temperature threshold, and adaptive control parameters are generated. This enables the motor to accelerate with maximum safe torque during startup, significantly improving the motor's power output capability and operational reliability while ensuring the thermal safety of the stator winding ends. Attached Figure Description

[0016] Figure 1 This is a structural diagram of a high-speed motor adaptive control system based on operating condition data, provided in one embodiment of this application. Figure 2 This is a flowchart illustrating a high-speed motor adaptive control method based on operating condition data, provided in one embodiment of this application. Detailed Implementation

[0017] The following description, in conjunction with the accompanying drawings, details a specific scheme for a high-speed motor adaptive control system based on operating condition data provided in this application.

[0018] Please see Figure 1 The diagram shows a structural diagram of a high-speed motor adaptive control system based on operating condition data provided in one embodiment of this application.

[0019] like Figure 1 As shown, the high-speed motor adaptive control system 10 based on operating condition data includes a data acquisition module 101, a loss determination module 102, a heat dissipation analysis module 103, a temperature prediction module 104, and an adaptive control module 105.

[0020] The data acquisition module 101 is used to acquire the stator winding end temperature and electrical excitation parameters of the high-speed motor at multiple times during startup.

[0021] Specifically, the data acquisition module 101 performs continuous sampling during the motor's acceleration from standstill to offline speed during startup. Because the stator winding current amplitude is large and the synchronous frequency rises rapidly during startup, the fundamental copper and iron losses increase sharply. At this time, the cooling oil pump has not yet established its rated operating pressure, and convective heat transfer is at its lowest level throughout the entire operation. The stator winding ends, as the area with the worst heat dissipation conditions, experience a rapid temperature rise. Therefore, the data acquisition module 101 needs to acquire synchronous sampling data that reflects this thermal dynamic process, namely, the stator winding end temperature and electrical excitation parameters.

[0022] Optionally, the electrical excitation parameters include the instantaneous values ​​of the multiphase current, the modulation ratio, and the synchronization frequency.

[0023] The instantaneous values ​​of the multiphase currents are provided by the current detection unit inside the controller. Based on the dual three-phase winding structure, the currents are divided into two groups of three-phase currents. Under the topology of neutral point isolation in each winding, the sum of the instantaneous values ​​of each group of three-phase currents is zero. The synchronization frequency is output by the phase-locked loop and represents the electrical angular frequency of the fundamental current in the stator winding. The modulation ratio is calculated from the real-time duty cycle of the space vector pulse width modulation, specifically the ratio of the amplitude of the fundamental phase voltage of the stator output by the controller to the amplitude of the maximum linearly modulated phase voltage.

[0024] It should be understood that within the linear modulation region of three-phase space vector pulse width modulation, the maximum undistorted line voltage amplitude of the inverter output is equal to the DC bus voltage. Since there is a fixed conversion relationship between phase voltage amplitude and line voltage amplitude in a three-phase system, with the coefficient being √3, the maximum linearly modulated phase voltage amplitude is determined based on the ratio of the DC bus voltage to √3; that is, the maximum linearly modulated phase voltage amplitude is equal to the DC bus voltage divided by √3.

[0025] It should be understood that the DC bus voltage is the bus voltage value sampled by the controller.

[0026] Optionally, the temperature at the end of the stator winding can be acquired by a temperature sensor embedded in the gap at the end of the stator winding. This temperature sensor can be a Pt1000 platinum resistance temperature sensor.

[0027] Optionally, the sampling frequency for synchronous sampling can be the controller's master control frequency.

[0028] The loss determination module 102 is used to determine the total motor loss at each time based on the electrical excitation parameters at each time.

[0029] It should be understood that total motor loss refers to all heat loss generated by the motor during operation at a given moment, and is the main heat source for end temperature rise.

[0030] In one optional implementation, the loss determination module 102 can decompose the total motor loss into fundamental copper loss, iron loss, and end-cap additional loss. Specifically, the loss determination module 102 is used to: determine the fundamental copper loss at each moment based on the synchronous frequency, the instantaneous value of the multiphase current at each moment, and the preset frequency AC resistance curve; determine the iron loss at each moment based on the synchronous frequency, the modulation ratio at each moment, and the preset frequency iron loss characteristic surface; determine the end-cap additional loss at each moment based on the fundamental copper loss at each moment and the preset frequency correction coefficient; and determine the total motor loss at each moment based on the fundamental copper loss, the iron loss, and the end-cap additional loss.

[0031] The preset frequency AC resistance curve refers to the offline calibrated curve that describes the relationship between the synchronous frequency and the single-phase AC resistance.

[0032] Optionally, the preset frequency AC resistance curve can be obtained by measuring or calculating the single-phase AC resistance value at multiple frequency points covering the motor operating frequency range under preset temperature conditions, and then fitting or interpolating each frequency point and its corresponding single-phase AC resistance value.

[0033] For example, the preset temperature condition can be the motor's rated operating temperature or the standard ambient temperature (e.g., 20 degrees Celsius); the motor's operating frequency range can cover from 0Hz to the highest synchronous frequency corresponding to the target offline speed; the number and distribution of frequency points can be set according to the interpolation accuracy requirements; the single-phase AC resistance value can be obtained by measuring the winding impedance with an impedance analyzer and then separating it, or by applying an AC excitation of a known frequency to the winding and measuring the voltage and current response and then calculating it.

[0034] Fundamental copper loss refers to the Joule heat loss generated by the fundamental current in the winding resistance.

[0035] In one alternative implementation, the effective value of each phase current at each moment can be determined based on the instantaneous value of the multiphase current at each moment; the single-phase AC resistance value at each moment can be obtained by interpolation from the AC resistance curve of the preset frequency based on the synchronization frequency at each moment; and the fundamental copper loss at each moment can be determined based on the effective value of each phase current and the single-phase AC resistance value at each moment.

[0036] It should be understood that the effective value of phase current is used to characterize the intensity of the thermal effect of the current.

[0037] Optionally, for each phase current, the instantaneous current value at multiple moments can be collected within the sliding time window to which each moment belongs, and the root mean square value of all instantaneous current values ​​within the sliding time window can be determined as the effective current value of that phase at each moment.

[0038] Optionally, the sliding time window to which each moment belongs is a window that traces back a fixed duration from that moment to the beginning. The preset length of the sliding time window can be the product of the temperature sensor response time constant and the sampling frequency, rounded up.

[0039] For example, assume the temperature sensor has a response time constant of 1500ms and a sampling frequency of 1ms. -1 Then the length of the sliding time window is 1500.

[0040] It should be understood that the single-phase AC resistance value refers to the single-phase winding resistance value after considering the skin effect at the corresponding synchronous frequency.

[0041] Optionally, the single-phase AC resistance value corresponding to the synchronization frequency at each moment in the preset frequency AC resistance curve can be determined as the single-phase AC resistance value at each moment.

[0042] Alternatively, the fundamental copper loss can be determined based on the relationship between the effective value of the phase current and the Joule loss of the AC resistance.

[0043] Optionally, the fundamental copper loss satisfies the following formula: in, Indicates time The fundamental copper loss, Indicates time The synchronization frequency, Indicates time Single-phase AC resistance, Indicates time The The effective value of the phase current. This formula represents the summation of the squared effective values ​​of the multiphase currents. It characterizes the total Joule losses in a multiphase winding due to the current-induced heating effect under single-phase AC resistance.

[0044] It should be understood that the preset frequency iron loss characteristic surface refers to the offline calibrated two-dimensional lookup table data, with the synchronous frequency and equivalent voltage utilization rate as inputs and iron loss as the output. Iron loss refers to the hysteresis loss and eddy current loss generated in the iron core by the alternating magnetic field.

[0045] Optionally, the preset frequency iron loss characteristic surface can be obtained by measuring or simulating iron loss values ​​at multiple operating points covering the motor's operating frequency range and voltage utilization range under the aforementioned preset temperature conditions, and then fitting or interpolating each operating point and its corresponding iron loss value. It should be understood that the voltage utilization range refers to the range of values ​​for the equivalent voltage utilization rate (i.e., modulation ratio) during motor operation. For example, this voltage utilization range can cover from 0 to the maximum linear modulation boundary of 1.0.

[0046] In one alternative implementation, the modulation ratio at each moment is determined as the equivalent voltage utilization rate at each moment; the iron loss at each moment is determined based on the synchronization frequency at each moment, the equivalent voltage utilization rate at each moment, and the iron loss characteristic surface at a preset frequency.

[0047] It should be understood that the equivalent voltage utilization rate is used to characterize the magnetic flux level of a motor.

[0048] Optionally, the synchronization frequency and equivalent voltage utilization rate at each time can be used as inputs to the preset frequency iron loss characteristic surface to obtain the iron loss at each time.

[0049] It should be understood that the preset frequency correction factor is used to characterize the proportion of additional end losses that vary with frequency. In the typical range of high current and low voltage utilization during startup, this factor can effectively reflect the additional heat contribution of the eddy current effect of the end structure as the frequency increases.

[0050] Among them, the additional end loss refers to the extra heat loss generated by the end leakage magnetic field and the eddy current effect of the structural components, in addition to the fundamental copper loss and iron loss.

[0051] Optionally, the preset frequency correction coefficient is obtained through offline finite element simulation calibration under the aforementioned preset temperature and operating conditions. Specifically, finite element simulation calculations are performed at multiple different synchronization frequency points: at each synchronization frequency point, the total motor loss considering the eddy current effect of the end structure and the total motor loss without considering the eddy current effect of the end structure are calculated; the difference between the total motor loss considering the eddy current effect of the end structure and the total motor loss without considering the eddy current effect of the end structure is taken as the additional end loss at that synchronization frequency point; the additional end loss is divided by the fundamental copper loss at the corresponding synchronization frequency point to obtain the frequency correction coefficient value at that synchronization frequency point; the frequency correction coefficient values ​​corresponding to multiple synchronization frequency points are curve-fitted to obtain the preset frequency correction coefficient that varies with the synchronization frequency.

[0052] Optionally, the preset operating conditions include multiple synchronous frequency points covering the motor's operating frequency range, as well as current amplitude and voltage utilization boundary conditions within the typical range of high current and low voltage utilization during startup.

[0053] For example, the current amplitude in the preset operating conditions can be set to the rated maximum torque current limit, and the voltage utilization rate can be set to the typical low voltage utilization rate during the start-up phase (e.g., 0.2 to 0.4). Multiple synchronous frequency points are selected within the motor operating frequency range, and the distribution density of synchronous frequency points should be increased in the frequency band where the end eddy current effect changes drastically.

[0054] Optionally, the fundamental copper loss at each time point can be multiplied by a preset frequency correction factor to obtain the additional end loss at each time point.

[0055] Optionally, the sum of the fundamental copper loss, iron loss, and additional end-point loss at each moment can be determined as the total motor loss at that moment.

[0056] In the above method, the total motor loss at each moment is decomposed into fundamental copper loss, iron loss, and end-component losses, making the calculation of total motor loss more refined and accurate. Calculating the fundamental copper loss by considering the AC resistance and effective value of the phase current in relation to frequency characteristics accurately reflects the winding heat loss under the skin effect; determining the iron loss by looking up the equivalent voltage utilization rate and synchronous frequency in a table reflects the coupling effect of magnetic flux level and frequency on core loss; and calculating the end-component losses by using a preset frequency correction coefficient reflects the often-neglected heat source of eddy current effects in the end-component structures.

[0057] The heat dissipation analysis module 103 is used to determine the current heat dissipation capacity and the trend of current heat dissipation capacity at the stator winding end based on the total motor loss at multiple times and the stator winding end temperature at multiple times.

[0058] It should be understood that the current heat dissipation capacity is used to characterize the change in stator winding end temperature caused by a unit loss increment under the current heat dissipation boundary conditions (i.e., the boundary constraints affecting the outward dissipation of heat from the ends at the current moment, including factors such as cooling oil flow rate, oil temperature, heat dissipation channel status, and ambient temperature). Its physical essence is dynamic thermal resistance. The current heat dissipation capacity change trend is used to characterize the evolution direction of heat dissipation capacity with the increase of total motor losses, that is, whether the heat dissipation conditions tend to improve or deteriorate.

[0059] It should be noted that the current time is the last of the multiple times of the above synchronous sampling.

[0060] Understandably, during startup, both the stator winding end temperature and the total motor loss monotonically increase over time. The thermal process at the motor end can be considered a dynamic system with the total motor loss as input and the stator winding end temperature as output, and its transient characteristics are determined by the equivalent heat dissipation capacity and heat capacity. When the heat dissipation conditions change with the cooling oil flow rate or oil temperature, the same increase in total motor loss corresponds to drastically different stator winding end temperature responses at different times. If we only observe the values ​​of the stator winding end temperature and the total motor loss in the time domain, we cannot directly know how much temperature rise will be caused by the unit loss increment at the current state point (i.e., the current total motor loss and stator winding end temperature), nor can we determine whether the heat dissipation conditions tend to improve or worsen.

[0061] Therefore, in this embodiment, the heat dissipation analysis module 103 uses the total motor loss as the independent variable and the stator winding end temperature as the dependent variable to construct a trajectory of state points moving over time at multiple moments within the mapping domain of loss and temperature. It then extracts the current heat dissipation capacity and the trend of its change from this trajectory. This trajectory is a broken line formed by sequentially connecting the state points corresponding to the total motor loss and the stator winding end temperature at multiple consecutive moments. The tangent direction of this trajectory directly reflects the temperature response law of the motor end thermal system to the loss input under the current heat dissipation boundary conditions. The tangent slope is used to characterize the amount of temperature change caused by the unit loss change under the current heat dissipation boundary conditions.

[0062] In one optional implementation, the heat dissipation analysis module 103 is specifically used to: use the total motor loss as the independent variable and the stator winding end temperature as the dependent variable to perform linear fitting on the total motor loss and stator winding end temperature within the sliding time window at each moment, and obtain the fitting line corresponding to each moment; use the slope of the fitting line at each moment as the heat dissipation capacity at each moment, and determine the heat dissipation capacity at the current moment as the current heat dissipation capacity; determine the average total motor loss within the sliding time window at each moment; and determine the trend of change of the current heat dissipation capacity based on the heat dissipation capacity at each moment and the average total motor loss at each moment.

[0063] Optionally, a monotonicity check can be performed on the state points within the sliding time window. If a local reverse jump occurs in an adjacent state point that does not conform to the overall trend of temperature increase and loss increase during the startup phase, the point is determined to be an abnormal disturbance point and is removed to ensure that the retained state points satisfy the thermodynamic causal direction consistency between the increase in loss and the increase in temperature.

[0064] Optionally, a weighted linear least squares method can be used for linear fitting. During the fitting process, the time decay weight of each state point is determined based on the time difference between each state point and the current time, as well as the sensor time constant, so that state points closer to the current time have a higher time decay weight. The slope of the resulting fitted line is then used as the current heat dissipation capacity, which can more accurately reflect the local heat dissipation boundary conditions at the current time and avoid interference from historical data far removed from the current time.

[0065] Optionally, the time decay weight satisfies the following formula: in, State point Time decay weight, Indicates the current moment. Representing state points The corresponding sampling time, Represents the current time and state point The absolute value of the time difference between the corresponding sampling times. This represents the sensor's time constant, used to eliminate the dimensions of the absolute value of the time difference. This represents an exponential function with the natural constant e as its base.

[0066] It is understandable that the slope of the fitted line has the physical meaning of the change in temperature at the end of the stator winding caused by a unit loss increment under the current heat dissipation boundary conditions. The smaller the slope, the stronger the heat dissipation capacity, that is, the smaller the temperature rise caused by the same loss increment.

[0067] It should be understood that the average total motor loss refers to the arithmetic mean of the total motor loss at all state points within the corresponding sliding time window, which is used to characterize the average heat load level within that sliding time window.

[0068] In one optional implementation, candidate values ​​for the heat dissipation capacity change trend at each moment can be determined based on the difference between the heat dissipation capacity at each moment and the heat dissipation capacity at the corresponding previous moment (hereinafter referred to as the heat dissipation capacity difference), and the difference between the average total motor loss at each moment and the average total motor loss at the corresponding previous moment (hereinafter referred to as the average loss difference). When a preset number of consecutive candidate values ​​for the heat dissipation capacity change trend meet a preset consistency condition, the candidate value for the heat dissipation capacity change trend at the current moment is determined as the current heat dissipation capacity change trend.

[0069] Optionally, the ratio between the difference in heat dissipation capacity at each time point and the difference in average loss can be determined as a candidate value for the trend of heat dissipation capacity change at each time point.

[0070] It should be noted that, in order to avoid the tiny slope measurement noise being amplified into a false trend of heat dissipation capacity change when the loss change is extremely small, when the average loss difference at a certain moment is less than the preset minimum loss difference threshold (e.g., 1 watt), the candidate value of the heat dissipation capacity change trend at that moment is determined to be invalid and will not participate in the subsequent confirmation process.

[0071] Optionally, the preset consistency conditions include: a preset number of consecutive candidate values ​​for the heat dissipation capacity change trend have the same positive or negative sign, and the ratio of the difference between their maximum and minimum values ​​to their absolute average value is less than a preset stability threshold (e.g., 10%).

[0072] For example, the preset quantity can be an empirical value of 3.

[0073] It is understandable that when a preset number of candidate values ​​for the heat dissipation capacity change trend meet the preset consistency condition, it means that the heat dissipation capacity change trend has maintained a stable sign direction and fluctuation amplitude over multiple consecutive moments, single-window noise interference has been effectively suppressed, and the candidate value for the heat dissipation capacity change trend at the current moment can truly reflect the evolution direction of the heat dissipation capacity as the operating conditions progress. At this time, the candidate value for the heat dissipation capacity change trend at the current moment can be confirmed as the current heat dissipation capacity change trend for subsequent temperature peak prediction.

[0074] When a preset number of candidate values ​​for the heat dissipation capacity change trend do not meet the preset consistency condition, it indicates that the direction of heat dissipation capacity change at multiple consecutive moments oscillates within multiple consecutive windows. The heat dissipation boundary conditions may be in a rapid transient period or be affected by strong interference, or the fluctuation amplitude of heat dissipation capacity is too large and the trend stability is insufficient. In this case, the current heat dissipation capacity change trend confirmed at the previous moment can be used until the preset consistency condition is met again in subsequent sliding time windows.

[0075] Optionally, if the current moment is the initial stage without a sliding time window, the current heat dissipation capacity is used as a constant value and the trend of heat dissipation capacity change is zero for deduction, until enough window data is accumulated and then a new judgment is made.

[0076] In the above method, the slope of the fitted straight line is extracted as the heat dissipation capacity through linear fitting. The candidate value of the heat dissipation capacity change trend is calculated by the ratio of the difference between the heat dissipation capacity and the average loss of adjacent sliding time windows. The candidate value is confirmed by a preset consistency condition, which effectively suppresses single-window noise and makes the extracted current heat dissipation capacity change trend truly reflect the evolution direction of heat dissipation conditions.

[0077] The temperature prediction module 104 is used to determine the predicted peak temperature based on the current heat dissipation capacity, the current heat dissipation capacity change trend, and the preset maximum loss target value.

[0078] It should be understood that the preset maximum loss target value refers to the total loss corresponding to the highest speed point that the motor is expected to reach under the current start-up control target, which represents the theoretical maximum thermal load that the motor needs to withstand during the start-up process.

[0079] The current startup control objectives include the target offline speed and the rated maximum torque current limit. The target offline speed is the final mechanical speed that the motor is expected to reach during startup, given by the upper-level control command. The rated maximum torque current limit is determined by the motor demagnetization limit or the driver hardware capability, representing the maximum allowable current limit during startup.

[0080] Optionally, the target offline speed and the rated maximum torque current limit can be obtained, and the corresponding highest synchronous frequency can be calculated by combining the number of pole pairs of the motor. Then, the rated maximum torque current limit is used as the current input for loss calculation, and the maximum value of the fundamental copper loss, the maximum value of the iron loss, and the maximum value of the end additional loss are determined according to the calculation method of the fundamental copper loss, the iron loss, and the end additional loss given above. The maximum value of the fundamental copper loss, the maximum value of the iron loss, and the maximum value of the end additional loss are accumulated to obtain the preset maximum loss target value.

[0081] The predicted peak temperature refers to the highest temperature that the stator winding end will reach when the total motor loss rises to the preset maximum loss target value, under the heat dissipation boundary conditions characterized by the current heat dissipation capacity and the trend of change in the current heat dissipation capacity.

[0082] Optionally, step-by-step iterative deduction can be performed based on the current heat dissipation capacity, the current trend of heat dissipation capacity change, and the current stator winding end temperature to predict the temperature when the preset maximum loss target value is reached, and this temperature is determined as the predicted temperature peak.

[0083] The adaptive control module 105 is used to determine the maximum allowable loss value based on the comparison result between the predicted peak temperature and the allowable end temperature threshold, and to generate adaptive control parameters based on the maximum allowable loss value.

[0084] It should be understood that the allowable end temperature threshold is the highest temperature limit that the stator winding end insulation material can safely withstand over a long period of time. It can be determined based on the insulation class and long-term temperature resistance characteristics of the stator winding end insulation material. In determining this threshold, firstly, the nominal long-term temperature resistance temperature is obtained based on the insulation material class selected for the motor (e.g., Class F or Class H insulation); then, considering the short-term overload characteristics and safety margin requirements during startup, a certain safety margin is reserved from the nominal long-term temperature resistance temperature to obtain the allowable end temperature threshold.

[0085] For example, assuming that the stator winding ends use Class H insulation material with a nominal long-term temperature resistance of 180°C, and considering a safety margin of 10°C to 20°C, the allowable temperature threshold at the ends can be set to 160°C to 170°C.

[0086] Optionally, when the predicted peak temperature is less than or equal to the allowable temperature threshold of the end, the preset maximum loss target value is determined as the maximum allowable loss value.

[0087] It is understandable that when the predicted temperature peak is less than or equal to the allowable temperature threshold of the end, it means that under the current heat dissipation conditions, accelerating to the target speed with the rated torque current will not exceed the allowable limit. There is no need to perform a limiting action, and the preset maximum loss target value can be directly used as the maximum allowable loss value. The corresponding adaptive control parameters maintain the rated setting value.

[0088] In one optional implementation, when the predicted peak temperature exceeds the allowable temperature threshold at the end, a binary search is performed with the current total motor loss value as the lower bound of the search interval and the preset maximum loss target value as the upper bound of the search interval. In each binary search, the loss value at the midpoint of the search interval is used as a candidate loss value, and step-by-step iterative deduction is performed based on the candidate loss value, the current heat dissipation capacity, and the trend of the current heat dissipation capacity to obtain the candidate predicted temperature value. The search interval is updated based on the comparison result between the candidate predicted temperature value and the allowable temperature threshold at the end, until the width of the search interval is less than the preset convergence tolerance. The loss value at the midpoint of the converged search interval is determined as the maximum allowable loss value.

[0089] It is understandable that when the predicted peak temperature is greater than the allowable temperature threshold at the end, it indicates that the current heat dissipation capacity is insufficient to support full torque acceleration, and the maximum allowable loss value corresponding to the thermal safety boundary must be solved.

[0090] It should be understood that in each binary search, when performing step-by-step iterative deduction, each iteration yields a loss value and a predicted temperature value. The iteration terminates when the loss value reaches or exceeds a candidate loss value. The predicted temperature value at the time of iteration termination is determined as the candidate predicted temperature value for this binary search, and the loss value at the time of iteration termination is determined as the candidate loss value for this binary search.

[0091] It should be noted that the step-by-step iteration method is the same as the step-by-step iteration method used by the temperature prediction module 104 to determine the predicted temperature peak value, and will not be described again here. Please refer to the following embodiments.

[0092] Understandably, if the candidate predicted temperature value is greater than the allowable end temperature threshold, it means that the candidate loss value obtained this time will still cause the end to overheat. In this case, the upper limit of the search interval will be narrowed to the candidate loss value obtained this time. If the candidate predicted temperature value is less than or equal to the allowable end temperature threshold, it means that the candidate loss value obtained this time is within the safe range. In this case, the lower limit of the search interval can be extended to the candidate loss value obtained this time.

[0093] Repeat the above process until the search interval width is less than the preset convergence tolerance. At this point, the search interval is narrow enough, and the midpoint of the interval can represent the maximum allowable loss value that makes the predicted temperature exactly equal to the allowable temperature threshold at the end.

[0094] After obtaining the maximum permissible loss value, the adaptive control module 105 can generate adaptive control parameters based on the maximum permissible loss value.

[0095] The adaptive control parameters may include torque current limiting value, cooling oil pump speed adjustment amount, and starting speed acceleration slope correction value.

[0096] Optionally, the torque current limit is determined as follows: the iron loss component corresponding to the target offline speed and the equivalent voltage utilization rate corresponding to the target offline speed is subtracted from the maximum allowable loss value to obtain the available margin of current-related losses; and the allowable fundamental copper loss is determined from the available margin of current-related losses based on the correspondence between end-additional losses and fundamental copper losses; then, the torque current limit is obtained by inverse calculation from the allowable fundamental copper losses based on the offline calibrated relationship between losses and torque current.

[0097] It should be understood that this available margin characterizes the upper limit of losses allowed by the thermal effects of electric current after deducting incompressible iron losses.

[0098] Optionally, to avoid the electromagnetic torque step caused by the sudden change in the torque current limit value, the adaptive control module 105 can also perform a smooth update of the torque current limit value according to the thermal time constant constraint, for example, by smoothing it using the exponential moving average method, and then output it to the current regulator for execution.

[0099] Optionally, the method for determining the feedforward adjustment amount of the cooling oil pump speed is as follows: the feedforward adjustment amount of the cooling oil pump speed is determined according to the reduction ratio of the maximum allowable loss value to the preset maximum loss target value.

[0100] The reduction ratio is used to characterize the tightness of the current thermal boundary relative to the theoretical maximum heat load. For example, the reduction ratio can be determined based on the ratio of the difference between the preset maximum loss target value and the maximum allowable loss value to the preset maximum loss target value. That is, the smaller the maximum allowable loss value, the larger the reduction ratio, indicating that the thermal boundary is tighter, and the feedforward adjustment of the cooling oil pump speed is increased accordingly to actively enhance the cooling flow and widen the heat dissipation boundary.

[0101] Optionally, the method for determining the closed-loop correction amount of the cooling oil pump speed is as follows: the difference between the preset maximum loss target value and the maximum allowable loss value is used as the error supply term, and the closed-loop correction amount of the cooling oil pump speed is calculated by the proportional controller.

[0102] It should be understood that the closed-loop correction is superimposed on the feedforward adjustment to form a cooling oil pump speed control command that coordinates feedforward and closed-loop, which realizes pre-compensation of heat dissipation capacity before the temperature at the end of the stator winding rises significantly. When the maximum allowable loss value is tightened to a lower level, the error supply term increases and the closed-loop correction is increased accordingly to avoid the thermal boundary margin being too small.

[0103] Optionally, the method for determining the start-up acceleration slope correction value is as follows: calculate the ratio of the maximum allowable loss value to the preset maximum loss target value, and use this ratio as the thermal allowable scaling factor. Multiply this thermal allowable scaling factor by the rated maximum acceleration slope to obtain the corrected start-up acceleration slope.

[0104] For example, when the maximum allowable loss value is greater than or equal to the preset maximum loss target value, the thermal allowable scaling factor is 1, and the start-up acceleration slope maintains the rated maximum acceleration slope; when the maximum allowable loss value is less than the preset maximum loss target value, the thermal allowable scaling factor is less than 1, and the start-up acceleration slope is linearly reduced according to the thermal allowable scaling factor.

[0105] Based on the description of the above embodiments, it should be understood that the high-speed motor adaptive control system 10 driven by operating condition data obtains the stator winding end temperature and electrical excitation parameters sampled synchronously at multiple moments during startup, providing a data foundation with strictly aligned timestamps for subsequent thermal analysis; by converting the electrical excitation parameters at each moment into the total motor loss at each moment, the input quantity for thermal drive is established; by extracting the current heat dissipation capacity and the current heat dissipation capacity change trend, the real-time heat dissipation status is accurately reflected; so that the current heat dissipation capacity, the current heat dissipation capacity change trend, and the preset maximum loss target value can be accurately predicted to obtain the peak temperature; finally, the maximum allowable loss value is determined and adaptive control parameters are generated based on the comparison result between the predicted peak temperature and the end allowable temperature threshold, so that the motor can accelerate with the maximum safe torque during startup, the thermal protection response is advanced and the amplitude limiting transition is smooth, significantly improving the power output capability and operational reliability.

[0106] Combination Figure 1 ,like Figure 2 As shown, when the temperature prediction module 104 determines the predicted peak temperature based on the current heat dissipation capacity, the current heat dissipation capacity change trend, and the preset maximum loss target value, it can be implemented by executing the following S201-S204.

[0107] S201. Determine the loss variation range within the sliding time window to which the current moment belongs based on the total motor loss at the start time and the total motor loss at the end time within the sliding time window to which the current moment belongs.

[0108] Specifically, the difference between the total motor loss at the start time and the total motor loss at the end time can be determined as the loss change amplitude, which reflects the change in the total motor loss under the current operating conditions.

[0109] S202. Determine the adaptive loss step size based on the loss variation magnitude and the number of moments contained within the sliding time window.

[0110] Optionally, the ratio between the magnitude of the loss change and the number of moments contained in the sliding time window can be determined as the adaptive loss step size. When the total motor loss increases rapidly, the step size automatically increases to improve the calculation efficiency; when the total motor loss changes slowly, the step size automatically decreases to ensure the calculation accuracy.

[0111] S203. Using the current total motor loss value as the initial loss value for iteration, the current stator winding end temperature as the initial temperature value for iteration, the current heat dissipation capacity as the initial heat dissipation capacity value for iteration, and the current heat dissipation capacity change trend as the iteration trend value, perform step iteration according to the adaptive loss step size to obtain the predicted loss value and predicted temperature value for each iteration step.

[0112] Specifically, in each iteration, the sum of the predicted loss value of the previous iteration step and the adaptive loss step size is determined as the predicted loss value of the current iteration step; the heat dissipation capacity value of the current iteration step is determined based on the adaptive loss step size, the iteration trend value, and the heat dissipation capacity value of the previous iteration step; the product of the adaptive loss step size and the heat dissipation capacity value of the current iteration step is determined as the temperature increment of the current iteration step; and the sum of the predicted temperature value of the previous iteration step and the temperature increment of the current iteration step is determined as the predicted temperature value of the current iteration step.

[0113] It should be noted that when an iteration step is the first iteration step, the initial loss value is determined as the predicted loss value of the previous iteration step, that is, the sum of the initial loss value and the adaptive loss step size is determined as the predicted loss value of the first iteration step. Similarly, the initial temperature value is determined as the predicted temperature value of the previous iteration step, that is, the sum of the initial temperature value and the temperature increment of the current iteration step is determined as the predicted temperature value of the first iteration step.

[0114] In one alternative implementation, the reliability coefficient of the current iteration step can be determined based on the current iteration step number; the product of the adaptive loss step size, the iteration trend value, and the reliability coefficient of the current iteration step can be determined as the heat dissipation capability correction amount of the current iteration step; and the sum of the heat dissipation capability value of the previous iteration step and the heat dissipation capability correction amount of the current iteration step can be determined as the heat dissipation capability value of the current iteration step.

[0115] It should be understood that the confidence coefficient should decrease monotonically with the current iteration step, representing the degree to which the extrapolation effectiveness used to characterize the trend of heat dissipation capacity decays as the number of iteration steps increases.

[0116] Optionally, the credibility coefficient satisfies the following formula: in, Indicates the current iteration step The credibility coefficient Indicates the current iteration step. Indicates the adaptive loss step size. The magnitude of the loss change, This is a preset constant used to prevent the denominator from being zero; for example, it can be 1 watt. Used to retrieve and The larger value in the formula ensures that the denominator will not be zero.

[0117] In this formula, as the current iteration number increases, the confidence coefficient decreases monotonically according to an exponential law. The less applicable the current heat dissipation capacity change trend is as a local rate of change, the lower the reliability of temperature extrapolation becomes.

[0118] It should be understood that the adaptive loss step size represents the incremental magnitude of the total motor loss in the current extrapolation step, the iterative trend value represents the current trend of heat dissipation capacity change, and the iterative trend value represents the direction and rate of evolution of heat dissipation capacity as loss increases. The reliability coefficient represents the degree of attenuation of the extrapolation effectiveness of the current heat dissipation capacity change trend in the current iteration step. The physical meaning of multiplying the adaptive loss step size, the iterative trend value, and the reliability coefficient is as follows: In the current iteration step, the adjustment amount of heat dissipation capacity calculated based on the heat dissipation capacity change trend, after being weighted and attenuated by the reliability coefficient, reflects the actual contribution of the extrapolation effectiveness of the current heat dissipation capacity change trend to the update of heat dissipation capacity.

[0119] Understandably, the sum of the heat dissipation capability value from the previous iteration and the heat dissipation capability correction amount from the current iteration represents the latest heat dissipation capability after curvature-driven correction.

[0120] It should be understood that multiplying the adaptive loss step size by the heat dissipation capacity value of the current iteration step means: under the current heat dissipation boundary conditions, the change in stator winding end temperature caused by an increase of one adaptive loss step size in the total motor loss. Adding the predicted temperature value of the previous iteration step to the temperature increment of the current iteration step yields the predicted temperature value for the current iteration step. This sum represents the new state of the stator winding end temperature after advancing one step along the starting trajectory.

[0121] Using the method provided in S203 above, the temperature prediction module 104 starts from the initial state at the current moment, driven by the total motor loss, and gradually recursively deduces the evolution trajectory of the stator winding end temperature. In each step of the deduction, the temperature increment, heat dissipation capacity value, predicted loss value and predicted temperature value are updated, and a reliable predicted temperature peak value is obtained under the premise of ensuring numerical stability.

[0122] S204. When the predicted loss value is greater than or equal to the preset maximum loss target value for the first time, the predicted temperature value of this iteration step is determined as the predicted temperature peak value.

[0123] Understandably, when the predicted loss value is initially greater than or equal to the preset maximum loss target value, it indicates that the total motor loss has reached the theoretical maximum thermal load boundary of the startup process in the simulation trajectory. At this time, the corresponding predicted temperature value is the temperature state that the motor end will reach when the total motor loss rises to the theoretical upper limit under the heat dissipation boundary conditions characterized by the current heat dissipation capacity and the current heat dissipation capacity change trend. Since the preset maximum loss target value is the theoretical end point of the startup process, and the loss no longer continues to increase, this predicted temperature value is the highest value of the stator winding end temperature during the entire startup process. It is determined as the predicted temperature peak value, which represents the extreme thermal state of the startup process under the current heat dissipation boundary.

[0124] The temperature prediction module 104 executes the methods provided in S201-S204 above, determining the adaptive loss step size by the loss change amplitude and the number of moments within the sliding time window, making the extrapolation step size consistent with the rhythm of the current thermal dynamic changes; by using the current total motor loss value, stator winding end temperature, current heat dissipation capacity, and current heat dissipation capacity change trend as the initial values ​​for step-by-step iterative extrapolation, the large-span analytical extrapolation is transformed into a controlled micro-step extrapolation, improving the accuracy of the iteration; by introducing a reliability coefficient that monotonically decreases with the number of iteration steps to dynamically constrain the extrapolation effectiveness of the current heat dissipation capacity change trend, the reliability of the prediction result decreases as the distance from the current moment increases, and reliable predicted temperature peaks can be obtained while ensuring numerical stability.

[0125] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0126] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A high-speed motor adaptive control system based on operating condition data, characterized in that, It includes a data acquisition module, a loss determination module, a heat dissipation analysis module, a temperature prediction module, and an adaptive control module: The data acquisition module is used to acquire the stator winding end temperature and electrical excitation parameters of the high-speed motor at multiple times during startup. The loss determination module is used to determine the total motor loss at each time based on the electrical excitation parameters at each time. The heat dissipation analysis module is used to determine the current heat dissipation capacity and the trend of change of the current heat dissipation capacity of the stator winding end based on the total motor loss at multiple times and the stator winding end temperature at multiple times. The temperature prediction module is used to determine the predicted peak temperature based on the current heat dissipation capacity, the trend of the current heat dissipation capacity, and the preset maximum loss target value. The adaptive control module is used to determine the maximum allowable loss value based on the comparison result between the predicted temperature peak and the allowable temperature threshold at the end, and to generate adaptive control parameters based on the maximum allowable loss value.

2. The high-speed motor adaptive control system based on operating condition data as described in claim 1, characterized in that, The electrical excitation parameters include the instantaneous value of the multiphase current, the modulation ratio, and the synchronization frequency. The loss determination module is specifically used for: The fundamental copper loss at each moment is determined based on the synchronization frequency at each moment, the instantaneous value of the multiphase current at each moment, and the AC resistance curve of the preset frequency. The iron loss at each moment is determined based on the synchronization frequency, modulation ratio, and iron loss characteristic surface at each moment. Based on the fundamental copper loss at each time point and the preset frequency correction coefficient, the additional end loss at each time point is determined. The preset frequency correction coefficient is used to characterize the proportion of additional end loss that varies with frequency. The total motor loss at each moment is determined based on the fundamental copper loss, the iron loss, and the additional end loss at each moment.

3. The high-speed motor adaptive control system based on operating condition data as described in claim 2, characterized in that, The loss determination module is also specifically used for: The effective value of each phase current is determined based on the instantaneous value of the multiphase current at each moment; Based on the synchronization frequency at each moment, the single-phase AC resistance value at each moment is obtained by interpolation from the AC resistance curve of the preset frequency. The fundamental copper loss at each moment is determined based on the effective value of the phase current and the single-phase AC resistance value at each moment.

4. The high-speed motor adaptive control system based on operating condition data as described in claim 2, characterized in that, The loss determination module is also specifically used for: The modulation ratio at each moment is determined as the equivalent voltage utilization rate at each moment; The iron loss at each moment is determined based on the synchronization frequency, the equivalent voltage utilization rate at each moment, and the iron loss characteristic surface at a preset frequency.

5. The high-speed motor adaptive control system based on operating condition data as described in claim 1, characterized in that, The heat dissipation analysis module is specifically used for: Using the total motor loss as the independent variable and the stator winding end temperature as the dependent variable, a linear fit is performed on the total motor loss and stator winding end temperature within the sliding time window at each moment to obtain the fitting straight line corresponding to each moment. The slope of the fitted line at each time point is taken as the heat dissipation capacity at each time point, and the heat dissipation capacity at the current time point is determined as the current heat dissipation capacity. Determine the average total motor loss within the sliding time window at each moment; The current trend of heat dissipation capacity change is determined based on the heat dissipation capacity at each time point and the average total motor loss at each time point.

6. The high-speed motor adaptive control system based on operating condition data as described in claim 5, characterized in that, The heat dissipation analysis module is also specifically used for: Based on the difference between the heat dissipation capacity at each moment and the corresponding heat dissipation capacity at the previous moment, and the difference between the average total motor loss at each moment and the average total motor loss at the previous moment, candidate values ​​for the heat dissipation capacity change trend at each moment are determined. When a preset number of consecutive candidate values ​​for the heat dissipation capacity change trend meet the preset consistency condition, the candidate value for the heat dissipation capacity change trend at the current moment is determined as the current heat dissipation capacity change trend.

7. The high-speed motor adaptive control system based on operating condition data as described in claim 1, characterized in that, The temperature prediction module is specifically used for: The loss variation range within the sliding time window to which the current moment belongs is determined based on the total motor loss at the start time and the total motor loss at the end time within the sliding time window to which the current moment belongs. The adaptive loss step size is determined based on the loss change magnitude and the number of moments contained within the sliding time window. The total motor loss value at the current moment is used as the initial loss value for iteration, the stator winding end temperature at the current moment is used as the initial temperature value for iteration, the current heat dissipation capacity is used as the initial heat dissipation capacity value for iteration, and the trend of the current heat dissipation capacity change is used as the iteration trend value. Step iteration is performed according to the adaptive loss step size to obtain the predicted loss value and predicted temperature value for each iteration step. When the predicted loss value is greater than or equal to the preset maximum loss target value for the first time, the predicted temperature value of this iteration step is determined as the predicted temperature peak value.

8. The high-speed motor adaptive control system based on operating condition data as described in claim 7, characterized in that, The temperature prediction module is also specifically used for: In each iteration, the sum of the predicted loss value of the previous iteration step and the adaptive loss step size is determined as the predicted loss value of the current iteration step. The heat dissipation capacity value of the current iteration step is determined based on the adaptive loss step size, the iteration trend value, and the heat dissipation capacity value of the previous iteration step. The product of the adaptive loss step size and the heat dissipation capacity value of the current iteration step is determined as the temperature increment of the current iteration step. The sum of the predicted temperature value from the previous iteration step and the temperature increment from the current iteration step is used to determine the predicted temperature value for the current iteration step.

9. The high-speed motor adaptive control system based on operating condition data as described in claim 8, characterized in that, The temperature prediction module is also specifically used for: The confidence coefficient of the current iteration step is determined based on the current iteration step number; The product of the adaptive loss step size, the iterative trend value, and the confidence coefficient of the current iteration step is determined as the heat dissipation capability correction amount for the current iteration step. The sum of the heat dissipation capacity value of the previous iteration step and the heat dissipation capacity correction amount of the current iteration step is determined as the heat dissipation capacity value of the current iteration step.

10. The high-speed motor adaptive control system based on operating condition data as described in claim 1, characterized in that, The adaptive control module is specifically used for: When the predicted peak temperature is less than or equal to the allowable end temperature threshold, the preset maximum loss target value is determined as the maximum allowable loss value. When the predicted peak temperature is greater than the allowable end temperature threshold, a binary search is performed with the current total motor loss value as the lower bound of the search interval and the preset maximum loss target value as the upper bound of the search interval. In each binary search, the loss value at the midpoint of the search interval is used as the candidate loss value, and step-by-step iterative deduction is performed based on the candidate loss value, the current heat dissipation capacity, and the trend of the current heat dissipation capacity to obtain the candidate predicted temperature value. Based on the comparison between the candidate predicted temperature value and the allowable temperature threshold at the end, the search interval is updated until the width of the search interval is less than the preset convergence tolerance. The loss value at the midpoint of the converged search interval is determined as the maximum allowable loss value.

Images