Motor overload fault detection method and system based on locked-rotor current characteristics

Abstract

The invention relates to the technical field of motor fault diagnosis, and discloses a motor overload fault detection method and system based on locked-rotor current characteristics. The method comprises the following steps: collecting a one-phase working current signal of a target driving motor; extracting three types of locked-rotor precursor features; calculating a locked-rotor risk score by using a nonlinear integral enhancement mechanism; judging whether to enter a motor overload state or not by adopting a multi-factor dynamic threshold joint judgment mechanism; compared with the prior art that locked-rotor judgment depends on single current amplitude abrupt change or an overcurrent threshold, and especially in a climbing scene that instantaneous locked-rotor current of a motor rises suddenly due to rapid load increase, stable perception and fault-tolerant response to overload faults cannot be achieved, the technical problem that in the prior art, locked-rotor judgment depends on single current amplitude abrupt change or overcurrent threshold judgment, and in the climbing scene that the instantaneous locked-rotor current of the motor rises suddenly is solved. Due to the fact that multi-dimensional precursor feature extraction and a nonlinear integral enhancement mechanism are combined, the problem of overload misjudgment or missed judgment is avoided, and the fault detection accuracy of the motor system under the complex working condition is improved.

Description

Technical Field

[0001] This invention relates to the field of motor fault diagnosis technology, and in particular to a method and system for detecting motor overload faults based on stall current characteristics. Background Technology

[0002] Currently, with the widespread application of electric tricycles in urban and rural logistics distribution, rural commuting, and micro-operations, the safety and stability of their drive systems have become increasingly important concerns. Especially under complex operating conditions such as slopes or fully loaded uphill runs, the drive motor is highly susceptible to extreme situations like sudden load surges or mechanical seizure, inducing a stall risk. Stall not only causes a sudden drop in motor output torque and vehicle slowdown, but can also lead to overcurrent in the drive unit, overheating of windings, and even burnout, severely impacting the reliability and lifespan of the entire vehicle. For example, when an electric tricycle is slowly starting on a slope or climbing at low speed, a sudden increase in the number of passengers, an overloaded cargo box, or a sudden increase in road friction can cause a sharp rise in the load on the motor. At this time, the drive current will rapidly exhibit a "nonlinear distortion segment," manifesting as steep peaks, transient spikes, and enhanced short-cycle disturbances—complex signal characteristics.

[0003] Existing protection mechanisms for motor stall can be mainly classified into the following categories: (1) Current threshold protection strategy: The stall state is determined by monitoring whether the current exceeds a preset threshold. Once the peak current exceeds the threshold, the drive is immediately cut off. Although this method is simple to implement, the current peak is often triggered only after the stall has occurred, lacking sufficient early warning capability. (2) Speed ​​judgment strategy: The stall situation is identified by comparing the deviation between the target speed and the actual speed. However, in low-speed climbing and other low-speed operating conditions, this method is prone to misjudging normal low speed as stall. (3) Back EMF model strategy: The motor load change is estimated based on back EMF modeling to predict the stall trend. However, in low-speed or even near-zero speed conditions, the back EMF is weak and easily affected by sampling error.

[0004] Therefore, there is an urgent need for a motor overload fault detection method based on stall current characteristics, which can still provide real-time, stable, and low-latency motor overload fault detection under conditions of sudden load increase and low-speed operation, so as to improve the control system's proactive response capability to extreme operating conditions. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the purpose of this invention is to propose a motor overload fault detection method based on stall current characteristics. This method aims to solve the technical problem that existing technologies rely on a single sudden change in current amplitude or overcurrent threshold to determine stall, especially in climbing scenarios where the instantaneous stall current of the motor rises sharply due to a rapid increase in load, making it impossible to achieve stable perception and fault-tolerant response to overload faults.

[0006] To solve the above technical problems, the present invention adopts the following technical solution: The present invention provides a method for detecting motor overload faults based on stall current characteristics. The motor overload fault detection method based on locked rotor current characteristics includes: Step S10: Acquire the operating current signal of one phase of the target drive motor at time t. Based on a single-phase operating current signal A velocity-state-aware dynamic limiting filter mechanism is used to perform precursor distortion suppression, and a preprocessed current sequence is output. ; Step S20: Based on the preprocessed current sequence A multi-criteria fusion mechanism is used to perform the task of extracting the pre-locked-rotation features, and three types of pre-locked-rotation features are output: current surge amplitude feature, current change rate feature, and local fluctuation entropy feature. Step S30: Based on the three types of pre-blocking warning features, a nonlinear integral enhancement mechanism is used to perform the task of calculating the blocking risk integral, and the blocking risk score is output. ; Step S40: Based on the congestion risk score A multi-factor dynamic threshold joint judgment mechanism is used to determine whether the target drive motor is in an overload state. Step S50: After determining that the target drive motor is in an overload state, the overload protection and tentative restart scheduling task is executed using a joint mechanism based on state locking and restart control, and the motor overload state locking signal is output.

[0007] Preferably, in step S10, the operating current signal of one phase of the target drive motor at time t is acquired. Based on a single-phase operating current signal A velocity-state-aware dynamic limiting filter mechanism is used to perform precursor distortion suppression, and a preprocessed current sequence is output. The steps specifically include: Step S101: Acquire the operating current signal of one phase of the target drive motor at time t with a sampling period of more than 1 ms. A limiting filter is used to filter the operating current signal of one phase. Perform digital denoising processing to output the intermediate current sequence. ; ;in, This is a preset amplitude limiting threshold function; This is the operating current signal of one phase at time t-1 corresponding to the previous sampling period; Step S102: Obtain motor speed state data ω(t), and based on the motor speed state data ω(t), use the adjacent abrupt change dynamic suppression method to process the intermediate current sequence. Dual-channel perturbation constraint enhancement processing is performed to output a preprocessed current sequence. .

[0008] Preferably, in step S102, the motor speed state data ω(t) is obtained, and the intermediate current sequence is processed based on the motor speed state data ω(t) using a dynamic suppression method based on adjacent abrupt changes. Dual-channel perturbation constraint enhancement processing is performed to output a preprocessed current sequence. The steps specifically include: Step S1021: Identify intermediate current sequences using a mutation identification method based on a combination of moving average and standard deviation. Local abnormal peak sections in; Step S1022: Calculate the velocity information entropy of the motor speed state data ω(t) using the sliding normalized probability distribution mapping method. Based on speed information entropy Set the dynamic adjustment factor of current suppression intensity , ; Step S1023: Dynamic adjustment factor based on current suppression intensity For intermediate current sequences The local abnormal peak sections in the output are perturbed using a gain scaling replacement method, and the final output is a preprocessed current sequence. .

[0009] Preferably, in step S20, based on the preprocessed current sequence The task of extracting stall precursor features using a multi-criteria fusion mechanism is to output three types of stall precursor features, including current surge amplitude features, current change rate features, and local fluctuation entropy features. The specific steps include: Step S201: Extraction of current surge amplitude: based on preprocessed current sequence Calculate the change in current between two adjacent sampling times as the current surge amplitude ΔI(t); if The current change exceeds the preset empirical peak value Then, the current surge amplitude ΔI(t) is taken as the effective feature value of the current surge, and all the effective feature values ​​of the current surge are extracted as the final current surge amplitude feature. Step S202: Calculate the rate of change of current v(t) per unit time based on the current surge amplitude ΔI(t), and set a threshold for the rate of change. ,when The current is determined to be in a "sudden rise followed by gradual change" state, and the rate of change of current v(t) in the "sudden rise followed by gradual change" state is extracted as the final rate of change of current feature. Step S203: Set a sliding window W, and use the sliding normalized probability distribution mapping method to statistically analyze the preprocessed current sequence within the sliding window W. Perturbation wave entropy If the perturbation wave entropy Less than the preset disturbance fluctuation entropy threshold If the current is in the "locked rotor stable section", then the disturbance fluctuation entropy in the "locked rotor stable section" state will be determined. This is extracted as the final local fluctuation entropy feature.

[0010] Preferably, in step S30, a nonlinear integral enhancement mechanism is used to perform the task of calculating the congestion risk integral based on the three types of congestion precursor characteristics, and the congestion risk score is output. The steps specifically include: Step S301: Construct the first risk index factor based on the amplitude characteristics of the current surge using amplitude normalization. A second risk index factor is constructed based on the characteristics of the current change rate using an inverse penalty method based on the amplitude of the current change rate. A third risk index factor is constructed based on the entropy characteristics of local fluctuations using an inverse penalty method based on entropy amplitude. ; Step S302: Based on the first risk index factor Second risk index factor and the third risk index factor Instantaneous risk factors are constructed using a nonlinear enhancement principle based on product fusion. ; Step S303: Based on instantaneous risk factors The dynamic forgetting amplification method is used to accumulate risk points and output a stall risk score. .

[0011] Preferably, in step S40, based on the stall risk score... The steps for determining whether a target drive motor is in an overload state using a multi-factor dynamic threshold joint judgment mechanism specifically include: Obtain the current change rate v(t) corresponding to the current change rate characteristic and the disturbance wave entropy corresponding to the local wave entropy characteristic. ; When the congestion risk score The score is greater than the preset congestion risk threshold and simultaneously meets the following conditions: If so, the target drive motor is determined to be in an overload state; where, The current rate of change convergence threshold. The threshold for convergence of perturbation wave entropy; When the congestion risk score Less than or equal to the preset blocking risk score threshold, and simultaneously satisfying If so, the target drive motor is determined to be in a non-motor overload state; In other cases, the target drive motor is determined to be in a suspected stall critical state.

[0012] Preferably, in step S50, after determining that the target drive motor is in an overload state, the step of executing overload protection and tentative restart scheduling tasks based on a joint mechanism of state locking and restart control, and outputting a motor overload state locking signal, specifically includes: Step S501: After determining that the target drive motor is in an overload state, immediately execute the shutdown operation of the target drive motor and terminate the PWM drive output of the target drive motor; Step S502: Preset and start the fault retry counter During the preset delay period Afterwards, if If so, the target drive motor will be restarted; where, Maximum number of restart attempts Step S503: During the preset delay period Afterwards, if If the motor is overloaded, an overload state lock signal will be output, and an alarm signal will be sent.

[0013] This invention also provides a motor overload fault detection system based on locked rotor current characteristics, comprising: The current acquisition and dynamic preprocessing module is used to acquire the one-phase operating current signal of the target drive motor at time t. Based on a single-phase operating current signal A velocity-state-aware dynamic limiting filter mechanism is used to perform precursor distortion suppression, and a preprocessed current sequence is output. ; A stall precursor feature extraction module is used to extract features based on preprocessed current sequences. A multi-criteria fusion mechanism is used to perform the task of extracting the pre-locked-rotation features, and three types of pre-locked-rotation features are output: current surge amplitude feature, current change rate feature, and local fluctuation entropy feature. The nonlinear integral enhancement calculation module is used to perform the task of calculating the congestion risk integral based on three types of congestion precursor characteristics using a nonlinear integral enhancement mechanism, and outputs the congestion risk score. ; The congestion assessment module is used to determine congestion risk based on the congestion risk score. A multi-factor dynamic threshold joint judgment mechanism is used to determine whether the target drive motor is in an overload state. The fault protection and tentative restart control module is used to execute overload protection and tentative restart scheduling tasks based on a joint mechanism of state locking and restart control after determining that the target drive motor is in an overload state, and outputs a motor overload state locking signal.

[0014] The present invention also provides a motor overload fault detection device based on locked rotor current characteristics, comprising: a memory, a processor, and a motor overload fault detection program based on locked rotor current characteristics stored in the memory and executable on the processor. When the motor overload fault detection program based on locked rotor current characteristics is executed by the processor, a motor overload fault detection method based on locked rotor current characteristics is implemented.

[0015] The present invention also provides a computer program product, including a motor overload fault detection program based on locked rotor current characteristics, wherein the motor overload fault detection program based on locked rotor current characteristics implements the motor overload fault detection method based on locked rotor current characteristics when executed by a processor.

[0016] The beneficial effects of this invention are as follows: By introducing a dynamic limiting filtering mechanism based on speed state perception and a multi-dimensional stall precursor feature extraction method (including current surge amplitude, current change rate and local fluctuation entropy), this invention can identify risk signals before stall is fully formed, significantly improving the sensitivity and real-time performance of stall precursor identification. Compared with the traditional detection method based on fixed current threshold, it has a stronger ability to resist operating condition fluctuation interference.

[0017] This invention employs a strategy combining nonlinear integral enhancement and multi-factor dynamic threshold determination. It performs adaptive overload determination based on risk scores and implements a phased fault response strategy by combining state locking and restart control mechanisms. This not only avoids erroneous shutdowns and recovery but also supports custom configuration of the maximum number of restart attempts and waiting time, thereby improving the stability of fault recovery and the fault tolerance of the motor control system. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the first embodiment of a motor overload fault detection method based on stall current characteristics according to the present invention.

[0020] Figure 2This is a schematic diagram of a device for a motor overload fault detection method based on stall current characteristics according to the present invention. Detailed Implementation

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

[0022] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of the motor overload fault detection method based on stall current characteristics of the present invention, which presents the first embodiment of the motor overload fault detection method based on stall current characteristics of the present invention.

[0023] In the first embodiment, the motor overload fault detection method based on locked rotor current characteristics includes: Step S10: Acquire the operating current signal of one phase of the target drive motor at time t. Based on a single-phase operating current signal A velocity-state-aware dynamic limiting filter mechanism is used to perform precursor distortion suppression, and a preprocessed current sequence is output. ; It should be noted that the "dynamic limiting and filtering mechanism based on speed state awareness" in this step refers to the process of adaptively limiting and filtering the acquired single-phase operating current signal by real-time monitoring of the target drive motor's speed state. This mechanism includes three core components: first, speed state identification, used to determine whether the motor is currently operating in a low-speed, stable, or high-speed range; second, dynamic limiting range adjustment, which adjusts the upper and lower limit thresholds of the current signal in real time according to different speed ranges, enabling the filter to suppress high-amplitude abrupt changes at low speeds and retain fast response characteristics at high speeds; and third, multi-stage smoothing filtering, which performs secondary smoothing on the limited current signal using methods such as moving average or median filtering to remove spike noise and pseudo-abrupt points. Through this mechanism, the output preprocessed current sequence can significantly reduce unstructured distortion interference while preserving the current trend change.

[0024] Understandably, this step effectively improves the resolvability and stability of motor current signals under complex operating environments. During low-speed climbing, frequent starts and stops, or heavy-load acceleration, the motor's current waveform is often accompanied by nonlinear distortion, harmonic interference, and electromagnetic noise. Directly using this for stall feature extraction would lead to distorted or even misjudged stall precursor identification. By introducing speed-state-aware dynamic amplitude-limiting filtering, this invention can adaptively adjust the filtering intensity for different speed stages, ensuring the integrity of the main current trend while improving the signal-to-noise ratio in the feature extraction stage, providing high-quality input for subsequent stall risk integration.

[0025] It should be understood that, compared to traditional methods using fixed threshold limiting or static filtering, this invention introduces speed state feedback during the filtering process, enabling dynamic adjustment of the filtering interval and freeing signal processing from reliance on empirically set values. Traditional limiting methods tend to mistakenly remove the initial current surge at low speeds as noise, while at high speeds they tend to retain excessive spikes, leading to increased signal fluctuations and failing to adapt to nonlinear current behavior under complex operating conditions. This invention forms an "adaptive limiting loop" through a speed-sensing mechanism, implementing differentiated suppression strategies in different speed ranges, thereby improving the fidelity of the precursor features of the current signal.

[0026] Step S20: Based on the preprocessed current sequence A multi-criteria fusion mechanism is used to perform the task of extracting the pre-locked-rotation features, and three types of pre-locked-rotation features are output: current surge amplitude feature, current change rate feature, and local fluctuation entropy feature. It should be noted that the "multi-criteria fusion mechanism" in this step refers to a comprehensive judgment process that involves parallel feature extraction and weighted fusion of different dimensions of motor current signal changes. This includes three key dimensions: first, the current surge amplitude feature, which reflects the severity of the current rise before stall occurs and captures the energy concentration effect in the early stage of stall by identifying transient amplitude changes in the current; second, the current change rate feature, which describes the speed and stability of the current rise and characterizes the transition behavior of the motor load from stable to abnormal by detecting the changing trend between adjacent sampling points; and third, the local fluctuation entropy feature, which measures the disorder and energy distribution complexity of the current waveform in a short period of time, thereby distinguishing between normal fluctuations and the disturbance concentration features induced by stall.

[0027] Understandably, by simultaneously extracting three types of stall precursor features within the same sampling window, early identification of stall trends from different physical dimensions can be achieved. The current surge amplitude feature provides an intuitive indicator of stall energy changes, the current change rate feature reflects the rapid change trend in the pre-stall stage, and the local fluctuation entropy feature reveals the convergence and aggregation patterns of current disturbances within the stall critical range. The combined application of these three features can effectively reduce misjudgments based on a single feature. Especially in low-speed, high-load, or interference-prone operating environments, the feature fusion mechanism can achieve stable identification of stall precursor signals, ensuring sufficient advance warning of risk trends.

[0028] For example, when an electric tricycle is going uphill at low speed, the motor current suddenly surges from 12A to 32A, and then the fluctuation gradually converges. In the traditional threshold method, this waveform would be ignored because it does not reach the fixed stall current threshold, and the protection would only be triggered after the stall actually occurs. However, in this invention, when the current surge amplitude reaches 1.8 times the normal value, the rate of change continuously decreases, and the fluctuation entropy is lower than the set threshold, a stall risk warning signal is output 200ms in advance. Experimental results show that, under the same operating conditions, this method improves the stall recognition advance by an average of about 25% and reduces the false judgment rate by about 40%, significantly improving the motor overload response speed and accuracy.

[0029] Step S30: Based on the three types of pre-blocking warning features, a nonlinear integral enhancement mechanism is used to perform the task of calculating the blocking risk integral, and the blocking risk score is output. ; It should be noted that the "nonlinear integral enhancement mechanism" in this step refers to a mechanism that, after obtaining the characteristics of current surge amplitude, current change rate, and local fluctuation entropy, continuously accumulates and dynamically amplifies the stall risk signal by establishing an integral model with adaptive weights and a forgetting factor. This mechanism includes three core components: first, establishing a multi-dimensional risk factor set to transform the three types of features into risk weight inputs under a unified scale; second, amplifying the risk weights through a nonlinear mapping function, exponentially enhancing the contribution of signals from sustained surges or convergent disturbances; and third, introducing a time forgetting parameter to automatically attenuate short-term anomalies, thereby ensuring that the integral process is sensitive to "persistent stall trends" and insensitive to "instantaneous false signals." Through this mechanism, the output stall risk score can accurately reflect the cumulative risk evolution trend of the motor's operating state.

[0030] Understandably, the technical advantage of this step lies in transforming traditional single-moment signal judgment into trend integral analysis over a continuous time domain, enabling the perception of the cumulative evolution of stall risk. By fusing three types of stall precursor features and performing nonlinear integration, the ability to detect "stall trend preceding stall event" can be achieved. When the current surge characteristic remains at a high amplitude, the rate of change gradually slows down, and the local fluctuation entropy decreases, the integral model rapidly amplifies the risk value, thereby outputting a high-confidence risk score before stall fully develops, providing a basis for subsequent overload protection and control decisions.

[0031] For example, in the case of an electric tricycle going uphill at low speed, when the vehicle load suddenly increases, the motor current rises continuously for a short period and then stabilizes, with the rate of change of current gradually decreasing and the entropy of disturbance fluctuations decreasing. Traditional overcurrent detection has not yet triggered an alarm at this stage, but the nonlinear integral enhancement mechanism of this invention gradually accumulates and amplifies this continuous signal with "high amplitude and low rate of change," so that the risk score exceeds the set threshold about 200 milliseconds before stalling occurs, thus triggering a stall warning in advance. Actual measurements show that this method improves the warning advance by about 23% compared to the traditional threshold method, reduces the false alarm rate by about 38%, and can output a high-confidence risk signal in time before stalling occurs, providing sufficient response time for the motor drive controller.

[0032] Step S40: Based on the congestion risk score A multi-factor dynamic threshold joint judgment mechanism is used to determine whether the target drive motor is in an overload state. It should be noted that the "multi-factor dynamic threshold joint judgment mechanism" in this step refers to the following: after calculating the stall risk score, it does not rely solely on a single integral value as the stall criterion, but simultaneously considers three criterion factors: the cumulative trend of the stall risk score, the convergence state of the current change rate, and the stability of the local fluctuation entropy. By dynamically weighting and fusing these three features, a judgment model with adaptive threshold adjustment capability is constructed. Specifically, the stall risk score represents the strength of the stall trend; the convergence of the current change rate is used to distinguish between normal starting torque and continuous stall load; and the convergence of the local fluctuation entropy reflects the stability of the current disturbance. These three factors together form a joint criterion, enabling robust stall judgment under different loads, speeds, and temperature rise conditions.

[0033] Understandably, by introducing dynamic threshold judgment logic, the identification of stall condition no longer relies on a fixed current threshold or a single risk integral value, but rather achieves adaptive adjustment based on real-time operating characteristics. When the current remains consistently high and the rate of change gradually converges, while the disturbance entropy shows a decreasing trend, it is determined that the motor has entered a stall critical state, thus outputting an overload alarm signal in advance. This multi-factor fusion judgment mechanism can effectively avoid false alarms caused by a single current surge, and also prevent the stall signal from lagging behind the actual physical state, ensuring the stability and real-time performance of the fault detection process.

[0034] For example, during the low-speed uphill operation of an electric tricycle, when the vehicle is fully loaded and encounters a sudden change in the friction coefficient of the slope, the motor current rapidly increases and then stabilizes within a short period, with the rate of change converging to a low value. Simultaneously, the current fluctuation entropy decreases to a stable range. At this point, the stall risk score remains high and continuously accumulates to exceed the dynamic threshold, indicating that the motor is in a stall state and an overload warning signal is issued. Actual test results show that, compared to traditional fixed-threshold detection, the method of this invention can identify stalls an average of approximately 200 milliseconds earlier, reducing the false positive rate by approximately 45%. It can reliably achieve early identification under load changes and low-speed conditions, effectively improving the operational safety and reliability of the motor drive.

[0035] Step S50: After determining that the target drive motor is in an overload state, the overload protection and tentative restart scheduling task is executed using a joint mechanism based on state locking and restart control, and the motor overload state locking signal is output.

[0036] It should be noted that the "state-locked and restart control joint mechanism" in this step refers to the following: when the target drive motor is determined to be in a stalled or severely overloaded state, the current operating state is first frozen and locked, and then a phased trial recovery process is executed according to the preset restart strategy. Specifically, it includes three steps: (1) State locking, that is, immediately interrupting the PWM drive output, cutting off the motor's main power path, and preventing the motor from continuously drawing excessive current, causing the windings to overheat or the driver to be damaged; (2) Restart control, after the fault is locked, the fault retry counter is started, the number of attempts is recorded, and the restart conditions are judged based on the motor temperature, bus voltage and stall risk score decay trend; (3) Trial scheduling, if the stall risk decreases and the restart conditions are met, a trial start will be executed by reducing the drive voltage and torque limit. If the stall risk signal is still detected, it will automatically revert and relock the state. This mechanism realizes the closed-loop control logic of "protection-recovery-reprotection".

[0037] It should be understood that, compared with existing motor control schemes that only have a "single-time shutdown protection" function, the state locking and restart control joint mechanism of this invention achieves adaptive scheduling and risk backtracking functions. Traditional methods often place the motor in a long-term shutdown state after stalling, or simply restart it periodically without distinguishing the cause of stalling, which can easily lead to repeated stalling, mechanical shock, or driver overheating. This invention introduces a dynamic assessment of stalling risk score and a maximum restart limit parameter into the restart control logic, which can autonomously decide whether to restart based on the risk decay trend. If there are multiple consecutive stalls, it automatically enters a long-term lockout state and triggers an alarm signal, thereby avoiding mechanical damage and extending the life of the drive module.

[0038] For example, taking the starting condition of an electric tricycle on a steep slope as an example, when the vehicle is overloaded and the slope is too steep, the motor stalls due to insufficient output torque. If the stall risk score is detected to be consistently higher than the dynamic threshold, a state lock is triggered, and the PWM output is immediately terminated, causing the drive current to rapidly decrease to a safe range. This stall event is recorded, and a 5-second cooling delay is initiated. When the motor temperature drops and the stall risk score falls to a safe range, a tentative restart is performed with 50% of the rated torque. If the motor current recovers smoothly and the fluctuation entropy remains low, normal drive is resumed; if stalling occurs again, the retry count is increased, and a second lockout is executed. If the restart attempt fails more than three times, automatic lockout is initiated, and an overload fault signal is uploaded to the main controller. Experimental verification shows that this mechanism can reduce the false lockout shutdown rate by approximately 40%, while shortening the average recovery time caused by short-term stalling by approximately 65%, significantly improving the reliability and self-healing ability of the motor under complex terrain conditions.

[0039] Example 2: Furthermore, the present invention provides a motor overload fault detection system based on locked-rotor current characteristics, employing a motor overload fault detection method based on locked-rotor current characteristics as described in the above embodiments, which can solve the technical problem of motor overload fault detection based on locked-rotor current characteristics. Compared with the prior art, the beneficial effects of the motor overload fault detection system based on locked-rotor current characteristics provided by the present invention are the same as those of the motor overload fault detection method based on locked-rotor current characteristics provided in the above embodiments, and other technical features of the motor overload fault detection system based on locked-rotor current characteristics are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0040] Example 3: This invention provides a motor overload fault detection device based on locked rotor current characteristics. Please refer to... Figure 2A motor overload fault detection device based on locked-rotor current characteristics includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to execute the motor overload fault detection method based on locked-rotor current characteristics described in Embodiment 1 above. The motor overload fault detection device based on locked-rotor current characteristics in this embodiment of the invention may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. This motor overload fault detection device based on locked-rotor current characteristics is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the invention. A motor overload fault detection device based on stall current characteristics may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the motor overload fault detection device based on stall current characteristics. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows a motor overload fault detection device based on locked-rotor current characteristics to communicate wirelessly or wiredly with other devices to exchange data. While the figure shows a motor overload fault detection device based on locked-rotor current characteristics with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0041] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for detecting motor overload faults based on locked-rotor current characteristics. The computer program product provided by this invention can solve the technical problem of detecting motor overload faults based on locked-rotor current characteristics. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the above-described method for detecting motor overload faults based on locked-rotor current characteristics, and will not be repeated here.

[0042] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0043] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0044] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for detecting an overload fault of an electric machine based on a stall current signature, characterized in that, The method comprises: Step S10: collect a phase working current signal of the target driving motor at time t , based on the phase working current signal Adopt a dynamic amplitude limiting filtering mechanism based on speed state perception to perform a precursor distortion suppression task, and output a preprocessed current sequence ; Step S20: based on the pre-processing current sequence The multiple criterion fusion mechanism is adopted to perform the stall precursor feature extraction task, and three types of stall precursor features are output, including the current sudden rise amplitude feature, the current change rate feature and the local fluctuation entropy feature. Step S30: Perform the stall risk integral calculation task based on the three types of stall precursor features using a nonlinear integral enhancement mechanism, output the stall risk score ; Step S40: determining the stall risk score Adopting a multi-factor dynamic threshold joint determination mechanism to determine whether the target driving motor is in an overload state. Step S50: after determining that the target driving motor is in the motor overload state, an overload protection and tentative restart scheduling task is performed based on a state locking and restart control combined mechanism, and an motor overload state locking signal is output.

2. A method of detecting motor overload faults based on stall current signature as claimed in claim 1, wherein, In step S10, a phase working current signal of the target driving motor at time t is collected , and a speed state sensing-based dynamic limiting filter mechanism is adopted to perform a precursor distortion suppression task based on the phase working current signal , and output a preprocessed current sequence , specifically including: Step S101: collect a phase working current signal of the target driving motor at time t with a sampling period higher than 1 ms , adopt a limiting filter to perform digital denoising processing on the phase working current signal , and output an intermediate current sequence ; ; wherein, is a preset limiting threshold function; is a phase working current signal corresponding to time t-1 of the previous sampling period; Step S102: Obtain motor speed state data ω(t), and use adjacent mutation dynamic suppression method based on the motor speed state data ω(t) to process the intermediate current sequence Perform double-channel disturbance constraint enhancement processing, and output the preprocessed current sequence .

3. A method of detecting motor overload faults based on stall current signatures as recited in claim 2, wherein, In step S102, the motor speed state data ω(t) is acquired, and the intermediate current sequence is suppressed based on the motor speed state data ω(t) using the adjacent mutation based dynamic suppression method The double-channel disturbance constraint enhancement processing is performed, and the preprocessed current sequence is output The step specifically includes: Step S1021: identify the local abnormal peak segment in the intermediate current sequence by using the mutation recognition method based on the combination of sliding mean and standard deviation ; Step S1022: calculate the speed information entropy of the motor speed state data ω(t) by using the sliding normalization probability distribution mapping method , based on the speed information entropy Set the current suppression strength dynamic adjustment factor , ; Step S1023: dynamically adjusting the factor based on the current suppression strength The intermediate current sequence The local abnormal peak segment in the intermediate current sequence is replaced by gain scaling to apply disturbance limitation, and finally outputs the preprocessed current sequence .

4. A method of detecting motor overload faults based on stall current signature as claimed in claim 1, wherein, In step S20, based on the pre-processed current sequence The step of adopting a multi-criterion fusion mechanism to perform a stall precursor feature extraction task and output three types of stall precursor features, including a current sudden rise amplitude feature, a current change rate feature, and a local fluctuation entropy feature, specifically includes: Step S201: extraction of current surge amplitude value: based on the pre-processed current sequence The current change amount between two adjacent sampling time points is calculated as the current surge amplitude value ΔI(t); if is greater than the preset current change empirical peak value , the current surge amplitude value ΔI(t) is taken as the current surge effective characteristic value, and all current surge effective characteristic values are taken as the final current surge amplitude value characteristics. Step S202: Calculate the current change rate v(t) in unit time based on the current surge amplitude ΔI(t), and set the change rate threshold , when , determine that the current is in the "slow change after surge" state, and extract the current change rate v(t) in the "slow change after surge" state as the final current change rate feature; Step S203: Set a sliding window W, and use the sliding normalized probability distribution mapping method to count the preprocessed current sequence in the sliding window W The disturbance fluctuation entropy If the disturbance fluctuation entropy is less than a preset disturbance fluctuation entropy threshold , it is determined that the current is in a "stall stable section" state, and the disturbance fluctuation entropy in the "stall stable section" state is extracted as the final local fluctuation entropy feature.

5. A method of detecting motor overload fault based on stall current signature as claimed in claim 1, wherein, In step S30, the stalling risk integral calculation task is performed based on the three types of stalling precursor features using a nonlinear integral enhancement mechanism, and a stalling risk score is output The steps specifically include: Step S301: constructing a first risk index factor based on the current surge amplitude characteristic using amplitude normalization ; constructing a second risk index factor based on the current rate of change characteristic using an inverse proportional penalty method based on the current rate of change amplitude ; constructing a third risk index factor based on the local fluctuation entropy characteristic using an inverse proportional penalty method based on the entropy amplitude ; Step S302: based on the first risk index factor , the second risk index factor , and the third risk index factor , a nonlinear enhancement principle based on product fusion is adopted to construct the instantaneous risk factor ; Step S303: Based on the instantaneous risk factor The risk score is output by using dynamic forgetting amplification method for risk score accumulation processing .

6. A method of detecting motor overload faults based on stall current signature as claimed in claim 4, wherein, In step S40, based on the stall risk score The step of determining whether the target driving motor is in an overload state by using the multi-factor dynamic threshold joint determination mechanism specifically includes: obtaining a current rate of change v(t) corresponding to the current rate of change feature and a perturbation fluctuation entropy corresponding to the local fluctuation entropy feature ; When the stall risk score is greater than a preset stall risk score threshold, and at the same time satisfies: , then it is determined that the target driving motor is in a motor overload state; wherein, is a current rate of change convergence threshold, is a disturbance fluctuation entropy convergence threshold; When the stall risk score is less than or equal to a preset stall risk score threshold, and at the same time , it is determined that the target drive motor is in a non-motor overload state; In other cases, it is determined that the target driving motor is in a suspected locked-rotor critical state.

7. A method of detecting motor overload faults based on stall current signature as claimed in claim 1, wherein, In step S50, after determining that the target driving motor is in the motor overload state, an overload protection and tentative restart scheduling task is performed based on a state locking and restart control combined mechanism, and an motor overload state locking signal is output. The step specifically comprises: Step S501: after determining that the target driving motor is in the motor overload state, immediately performing a shutdown operation of the target driving motor, and terminating PWM driving output of the target driving motor; Step S502: Preset and start the fault retry counter During the preset delay period Afterwards, if If so, the target drive motor will be restarted; where, Maximum number of restart attempts Step S503: In the preset delay period After that, if The motor overload state locking signal is output, and an alarm signal is sent.

8. A system for detecting motor overload fault based on stall current signature, applied to the method for detecting motor overload fault based on stall current signature according to any one of claims 1 to 7, characterized in that, The motor overload fault detection system based on locked-rotor current characteristics comprises: The current acquisition and dynamic preprocessing module is used for acquiring a phase working current signal of the target driving motor at time t , based on the phase working current signal Adopting a dynamic amplitude limiting filtering mechanism based on speed state perception to perform a precursor distortion suppression task, and outputting a preprocessed current sequence ; A stall precursor feature extraction module is configured to extract stall precursor features based on the preprocessed current sequence The stall precursor feature extraction task is performed by using a multi-criterion fusion mechanism, and three types of stall precursor features are output, including a current sudden rise amplitude feature, a current change rate feature, and a local fluctuation entropy feature. The nonlinear integral enhancement computing module is configured to perform a stall risk integral calculation task based on the three types of stall precursor features using a nonlinear integral enhancement mechanism, and output a stall risk score ; A stall determination module is configured to determine a stall risk score based on the motor current and the motor speed A multi-factor dynamic threshold joint determination mechanism is adopted to determine whether the target driving motor is in an overload state. A fault protection and tentative restart control module is configured to, after determining that the target driving motor is in the motor overload state, perform an overload protection and tentative restart scheduling task based on a state locking and restart control combined mechanism, and output an motor overload state locking signal.

9. A stall current signature based motor overload fault detection apparatus, characterized by, The motor overload fault detection device based on locked-rotor current characteristics comprises a memory, a processor, and a motor overload fault detection program based on locked-rotor current characteristics stored on the memory and executable on the processor. When the motor overload fault detection program based on locked-rotor current characteristics is executed by the processor, the motor overload fault detection method based on locked-rotor current characteristics in any one of claims 1 to 7 is implemented.

10. A computer program product, characterised in that, The computer program product comprises a motor overload fault detection program based on locked-rotor current characteristics. When the motor overload fault detection program based on locked-rotor current characteristics is executed by the processor, the motor overload fault detection method based on locked-rotor current characteristics in any one of claims 1 to 7 is implemented.

Images