A self-diagnostic system and device for stall faults in handheld vacuum cleaners

Abstract

This invention relates to the field of household appliance fault diagnosis technology, and provides a self-diagnostic system and device for stall faults in handheld vacuum cleaners. The self-diagnostic system includes: a core electrical parameter acquisition and dimensionless per-unit conversion module for real-time acquisition of the bus voltage, phase current, duty cycle, and motor speed of the brushless DC motor of the handheld vacuum cleaner, and dimensionless per-unit conversion of the above physical quantities using the motor's rated reference parameters, mapping them to per-unit values; and a pneumatic load index sensorless observation module for calculating an independent pneumatic load index unaffected by speed fluctuations. This invention, by constructing a two-dimensional orthogonal coordinate system of the amplitude mean and high-frequency variance fluctuation characteristics of the pure mechanical jamming torque, can accurately separate three operating conditions: "rigid foreign object jamming," "carpet flexible legal fluctuation," and "air duct pneumatic blockage."

Description

Technical Field

[0001] This invention belongs to the field of household appliance fault diagnosis technology, and in particular relates to a self-diagnosis system for stall faults in handheld vacuum cleaners. Background Technology

[0002] Handheld vacuum cleaners, as a core tool for modern household cleaning, rely on a brushless DC motor-driven centrifugal fan system for their portability and efficiency. In real-world use, vacuum cleaners frequently encounter various complex operating conditions leading to abnormal loads. These abnormalities are typically accompanied by increased motor current and decreased motor speed, prompting traditional stall protection mechanisms to trigger a shutdown command based solely on whether the absolute value of the current or speed exceeds a fixed threshold. However, this simplistic threshold-based method has fundamental flaws.

[0003] When a vacuum cleaner is running on a long-pile carpet, the periodic resistance fluctuations caused by the carpet fibers can induce instantaneous peak current. If the protection system responds directly to such fluctuations, it will lead to frequent false shutdowns, seriously interfering with the user's cleaning process. Conversely, if the threshold is relaxed to avoid false alarms, it may lose its ability to provide timely protection when a real rigid jam occurs, thus creating a risk of equipment damage.

[0004] More notably, when the duct is blocked, the centrifugal fan's air intake decreases sharply, the motor load is reduced, and the current may show a downward trend, but the vacuum cleaner has completely lost its vacuuming function. The existing stall detection logic only focuses on the overcurrent state and cannot identify this type of "underload" fault. Users often only notice the problem after the cleaning effect has deteriorated significantly, which not only wastes electricity but may also cause overheating risks due to the continuous inefficient operation of the motor.

[0005] Existing technologies are inadequate in distinguishing between fault types such as rigid jamming, carpet elastic fluctuations, and airflow blockage, making it difficult to meet users' urgent needs for vacuum cleaner reliability and intelligent diagnostics.

[0006] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0007] The purpose of this invention is to provide a self-diagnostic system for stall faults in handheld vacuum cleaners, aiming to solve the above-mentioned problems.

[0008] This invention is implemented as follows: a self-diagnostic system for stall faults in a handheld vacuum cleaner, comprising:

[0009] Core electrical parameter acquisition and dimensionless per-unit conversion module: used to acquire the bus voltage, phase current, duty cycle and motor speed of the brushless DC motor of the handheld vacuum cleaner in real time, and to perform dimensionless per-unit conversion on the above physical quantities using the rated reference parameters of the motor, mapping them to per-unit values.

[0010] Aerodynamic load index non-intrusive observation module: used to establish an aerodynamic observer based on the centrifugal fan similarity law, remove the inertial component from the per-unit value of electromagnetic torque, divide it by the cube of the per-unit value of speed, and calculate the independent aerodynamic load index that is not affected by speed fluctuations, so as to characterize the true flow area of ​​the current air duct.

[0011] Axial negative pressure friction cross compensation module: used to establish aerodynamic-mechanical cross coupling equation based on the rotor axial thrust effect caused by fluid mechanics; and dynamically calculate the per-unit value of the bearing additional friction torque caused by air duct blockage using the drop difference value of the aerodynamic load index.

[0012] Pure mechanical jamming torque extraction module: used to accurately deduct the acceleration inertia torque, aerodynamic load torque and the bearing additional friction torque per unit value from the total electromagnetic torque per unit value of the motor, and extract the pure mechanical jamming torque caused only by the entrainment of external physical foreign objects;

[0013] Transient variance fluctuation characteristic calculation module: used to establish a sliding time window and calculate the discrete high-frequency variance fluctuation characteristics of the pure mechanical jamming torque in the time domain;

[0014] Two-dimensional orthogonal decoupling and fault diagnosis module: used to construct a two-dimensional orthogonal coordinate system by combining the amplitude mean of the pure mechanical jamming torque with the high-frequency variance fluctuation characteristics, accurately separate the fault boundaries of "rigid foreign object jamming", "carpet flexible legal fluctuation" and "air duct aerodynamic blockage", and output classification intervention instructions.

[0015] A further technical solution is that the core electrical parameter acquisition and dimensionless per-unit conversion module performs dimensionless per-unit conversion as follows: the real-time acquired bus voltage, phase current, and motor speed are divided by the corresponding rated bus voltage, rated phase current, and rated maximum speed to obtain their respective per-unit values; through this processing, electrical parameters of different dimensions are uniformly compressed into a proportional scale to prevent data overflow faults from occurring in the microcontroller during multiplication and division operations.

[0016] A further technical solution is that the aerodynamic load index non-sensory observation module calculates the aerodynamic load index in the following specific way:

[0017] First, calculate the inverter efficiency coefficient by multiplying the bus voltage per unit value, the phase current per unit value, and the duty cycle by the product. Then, subtract the square of the phase current per unit value multiplied by the per-unit stator phase resistance, and then subtract the per-unit rotor total moment of inertia multiplied by the derivative of the speed per unit value with respect to time, and then multiply by the speed per unit value.

[0018] Then divide the above result by the cube of the per-unit speed value to obtain the aerodynamic load index at the current moment.

[0019] A further technical solution is that the cross-coupling equation for the dynamic calculation of the per-unit value of the bearing's additional frictional torque by the axial negative pressure frictional force cross-compensation module is specifically as follows:

[0020] The per-unit value of the base static friction torque under normal pressure free state is added to the square of the rotor axial force-friction force cross-coupling conversion constant multiplied by the positive part of the difference between the reference aerodynamic load index and the current aerodynamic load index, to obtain the per-unit value of the bearing additional friction torque including the vacuum degree additional effect.

[0021] A further technical solution is that the specific formula for the pure mechanical jamming torque extraction module to extract the pure mechanical jamming torque is as follows:

[0022] First, calculate the per-unit value of the total electromagnetic torque of the motor. This value is equal to the inverter efficiency coefficient multiplied by the per-unit value of the bus voltage, the per-unit value of the phase current, and the duty cycle in sequence. Subtract the square of the per-unit value of the phase current multiplied by the per-unit stator phase resistance from the result, and then divide the difference by the per-unit value of the speed.

[0023] Then, subtract the acceleration inertia torque, aerodynamic load torque, and bearing additional friction torque per unit value sequentially from the total electromagnetic torque per unit value. The acceleration inertia torque is equal to the per-unit rotor total moment of inertia multiplied by the differential of the per-unit speed value, and the aerodynamic load torque is equal to the aerodynamic load index multiplied by the square of the per-unit speed value. Finally, the pure mechanical jamming torque per unit value, which decouples the duct vacuum degree from the motor inertial interference, is obtained.

[0024] A further technical solution is that the transient variance fluctuation characteristic calculation module calculates the high-frequency variance fluctuation characteristics in the following specific way:

[0025] Establish a sliding time window of length N, calculate the mean amplitude of the pure mechanical jamming torque within the window; then calculate the square of the difference between the torque value and the mean at all times within the window, sum these squared values, and divide by the window length N to obtain the high-frequency variance fluctuation characteristics.

[0026] A further technical solution is that the logic for performing two-dimensional orthogonal decoupling and fault diagnosis by the two-dimensional orthogonal decoupling and fault diagnosis module specifically includes:

[0027] When the average amplitude of the pure mechanical jamming torque is greater than the safety warning threshold and the high-frequency variance fluctuation characteristic is less than the first variance constant, it is determined to be rigid mechanical jamming, and an electronic braking command is output.

[0028] When the average amplitude is greater than the safety warning threshold and the high-frequency variance fluctuation characteristic is greater than the second variance constant, it is judged as a carpet flexible legal fluctuation and the alarm action is blocked.

[0029] When the average amplitude approaches zero and the aerodynamic load index is lower than the preset ventilation threshold, it is determined that the air duct is aerodynamically blocked, and a frequency reduction and filter cleaning prompt command is output.

[0030] The first variance constant is smaller than the second variance constant.

[0031] A self-diagnostic device for stall faults in a handheld vacuum cleaner, employing any of the systems described above, comprising:

[0032] The vacuum cleaner body, and the brushless DC motor, air duct centrifugal impeller, inverter drive circuit and sensorless sampling circuit installed inside the vacuum cleaner body;

[0033] The handheld vacuum cleaner stall fault self-diagnosis system is burned into the microcontroller in the form of firmware code;

[0034] The microcontroller reads the electrical signal of the inverter drive circuit through the sensorless sampling circuit, uses the built-in stall fault self-diagnosis system to output the current fault classification command in real time, and drives the inverter drive circuit to perform physical protection actions such as electronic brake stop, low duty cycle reverse or frequency reduction operation.

[0035] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0036] This invention constructs a two-dimensional orthogonal coordinate system that integrates the amplitude mean and high-frequency variance fluctuation characteristics of pure mechanical jamming torque. This system can accurately separate three operating conditions: "rigid foreign object jamming," "carpet flexible legal fluctuation," and "air duct aerodynamic blockage." When carpet elastic fluctuation is detected, the alarm is automatically suppressed to avoid frequent false shutdowns; while in the case of rigid jamming, an electronic braking command is immediately output to ensure the safety of the equipment and the user.

[0037] Based on the axial thrust effect of the rotor in hydrodynamics, an aerodynamic-mechanical cross-coupling equation is established. The per-unit value of the additional bearing frictional torque caused by duct blockage is dynamically calculated using the drop difference of the aerodynamic load index. This compensation mechanism accurately subtracts the time-varying frictional component from the total electromagnetic torque, significantly improving the accuracy of extracting pure mechanical jamming torque and the reliability of fault diagnosis.

[0038] The aerodynamic load index non-sensory observation module characterizes the actual flow area of ​​the air duct in real time. When the pure mechanical jamming torque approaches zero and the aerodynamic load index is lower than the ventilation threshold, it is determined that the filter or extension tube is completely blocked, and the output frequency is reduced and a cleaning prompt is issued. This solves the blind spot problem of traditional overcurrent protection failing to detect air duct blockage (current drop), avoiding energy waste and motor overheating.

[0039] By dividing key electrical parameters such as bus voltage, phase current, and speed by the rated reference value, and mapping them to dimensionless per-unit values, the data range is effectively compressed, preventing data overflow faults when the microcontroller executes complex dynamic equations such as power multiplication and division. This improves the robustness and engineering practicality of the algorithm on low-cost microcontrollers. Attached Figure Description

[0040] Figure 1 This is a diagram of the overall functional module architecture of the system;

[0041] Figure 2 Logic diagram for extracting jamming torque and calculating variance features;

[0042] Figure 3 This is a two-dimensional orthogonal decoupling and classification diagnosis decision diagram. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0045] like Figure 1 As shown, a self-diagnostic system for stall faults in a handheld vacuum cleaner, according to an embodiment of the present invention, includes:

[0046] Core electrical parameter acquisition and dimensionless per-unit conversion module: used to acquire the bus voltage, phase current, duty cycle and motor speed of the brushless DC motor of the handheld vacuum cleaner in real time, and to perform dimensionless per-unit conversion on the above physical quantities using the rated reference parameters of the motor, mapping them to per-unit values.

[0047] Bus voltage, phase current, duty cycle, and motor speed are key electrical parameters describing the operating state of a brushless DC motor. Bus voltage and phase current reflect the motor's power input, the duty cycle controls the output of the motor drive circuit, and the motor speed directly relates to the fan's suction force and the motor load. Dimensionless per-unit scaling is a data preprocessing technique that aims to convert physical quantities with different dimensions and numerical ranges into dimensionless proportional values ​​(i.e., per-unit values). By comparing actual measured values ​​with the motor's rated reference parameters, the data range can be effectively compressed, preventing data overflow in subsequent complex calculations and improving the robustness of the algorithm.

[0048] Aerodynamic load index non-intrusive observation module: used to establish an aerodynamic observer based on the centrifugal fan similarity law, remove the inertial component from the per-unit value of electromagnetic torque, divide it by the cube of the per-unit value of speed, and calculate the independent aerodynamic load index that is not affected by speed fluctuations, so as to characterize the true flow area of ​​the current air duct.

[0049] Among them, the aerodynamic load index is an independent parameter characterizing the actual flow area of ​​the current air duct. It is achieved by establishing an aerodynamic observer based on the centrifugal fan similarity law, extracting the aerodynamically related components from the electromagnetic torque of the motor, and normalizing the speed. Thus, it can independently reflect the unobstructedness of the air duct and is not affected by fluctuations in motor speed.

[0050] Axial negative pressure friction cross compensation module: used to establish aerodynamic-mechanical cross coupling equation based on the rotor axial thrust effect caused by fluid mechanics; and dynamically calculate the per-unit value of the bearing additional friction torque caused by air duct blockage using the drop difference value of the aerodynamic load index.

[0051] The axial negative pressure friction cross-compensation module is used to handle the coupling relationship between hydrodynamic effects and mechanical friction. When the vacuum cleaner is working, the axial negative pressure generated by the centrifugal fan applies a thrust to the rotor bearing, thereby changing the bearing's friction torque. This module establishes a pneumatic-mechanical cross-coupling equation to dynamically calculate and compensate for the additional bearing friction torque caused by duct blockage, thereby improving the accuracy of mechanical jamming torque extraction.

[0052] The pure mechanical jamming torque extraction module is used to accurately subtract the acceleration inertia torque, aerodynamic load torque, and bearing additional friction torque per unit from the total electromagnetic torque per unit value of the motor, extracting the pure mechanical jamming torque caused solely by the entrainment of external physical foreign objects. This pure mechanical jamming torque refers to the mechanical resistance torque caused solely by external physical foreign objects (such as cables, hair, etc.) entrained in the floor brush or air duct. By accurately subtracting the acceleration inertia torque, aerodynamic load torque, and bearing additional friction torque from the total electromagnetic torque of the motor, this pure mechanical jamming component can be extracted, thereby eliminating interference from other non-faulty load fluctuations.

[0053] The transient variance fluctuation characteristic calculation module is used to establish a sliding time window and calculate the discrete high-frequency variance fluctuation characteristics of the pure mechanical jamming torque in the time domain. Specifically, this module analyzes the dynamic changes of the pure mechanical jamming torque in the time domain. By establishing a sliding time window and calculating the discrete high-frequency variance of the torque signal within that window, the essential differences in the time scale between different types of loads (e.g., rigid jamming and flexible fluctuation) can be captured.

[0054] Two-dimensional orthogonal decoupling and fault diagnosis module: used to construct a two-dimensional orthogonal coordinate system by combining the amplitude mean of the pure mechanical jamming torque with the high-frequency variance fluctuation characteristics, accurately separate the fault boundaries of "rigid foreign object jamming", "carpet flexible legal fluctuation" and "air duct aerodynamic blockage", and output classification intervention instructions.

[0055] The two-dimensional orthogonal decoupling and fault diagnosis module is the final decision-making unit of the system. It constructs a two-dimensional orthogonal coordinate system by combining the amplitude mean of the pure mechanical jamming torque with the high-frequency variance fluctuation characteristics. In this coordinate system, different types of faults (such as rigid foreign object jamming, carpet flexible legal fluctuation, and air duct aerodynamic blockage) will be mapped to different regions, thereby achieving accurate separation of fault boundaries and output of classification intervention commands.

[0056] In this embodiment, firstly, the core electrical parameter acquisition and dimensionless per-unit scaling module continuously acquires the bus voltage, phase current, duty cycle, and motor speed of the brushless DC motor of the vacuum cleaner during operation. For example, through built-in sensors and sensorless observers, the system can collect data every 1 millisecond. Subsequently, these raw physical quantities are dimensionlessly scaled using the motor's rated reference parameters. For example, the actual speed is divided by the rated maximum speed, and the actual current is divided by the rated phase current, thereby uniformly mapping all data to a per-unit value range of 0 to 1, ensuring the stability and accuracy of subsequent calculations.

[0057] Next, the aerodynamic load index non-intrusive observation module uses this standardized data to establish an aerodynamic observer based on the centrifugal fan similarity law. This observer calculates an independent aerodynamic load index in real time, unaffected by speed fluctuations. For example, when the vacuum cleaner is operating normally on an open surface, the airflow is unobstructed, and the aerodynamic load index remains near a high baseline value, indicating a large airflow area. However, when the vacuum cleaner sucks in a large amount of dust, causing the filter to become clogged, the airflow area decreases, and the aerodynamic load index drops accordingly.

[0058] Simultaneously, the axial negative pressure friction cross-compensation module establishes an aerodynamic-mechanical cross-coupling equation based on the rotor axial thrust effect caused by fluid mechanics. This module dynamically calculates the per-unit value of the additional bearing friction torque caused by duct blockage using the aforementioned drop in the aerodynamic load index. For example, when the aerodynamic load index begins to decrease from the baseline value, the system calculates the additional torque caused by the change in negative pressure due to bearing friction, based on a preset coupling relationship, and uses this as a compensation term.

[0059] Subsequently, the pure mechanical jamming torque extraction module precisely deducts the acceleration inertia torque, aerodynamic load torque, and the aforementioned bearing additional friction torque per unit value from the total electromagnetic torque per unit value of the motor. For example, when the motor speed changes, inertia torque is generated; when the air duct is unobstructed, aerodynamic load torque is generated; when the air duct is blocked, additional friction torque is generated. By deducting these non-jamming factors one by one, the system can extract the pure mechanical jamming torque caused solely by the entrapment of external physical foreign objects.

[0060] Next, the transient variance fluctuation characteristic calculation module establishes a sliding time window, for example, a time window containing the most recent 100 sampling points. Within this window, the system calculates the discrete high-frequency variance fluctuation characteristics of the aforementioned pure mechanical locking torque in the time domain. For example, if the floor brush gets caught in a cable, the amplitude of the pure mechanical locking torque will increase significantly, but its fluctuation (variance) will be small because the cable provides rigid resistance. However, if the floor brush is working on a long-pile carpet, the amplitude of the pure mechanical locking torque may also increase, but its fluctuation (variance) will be very large because the carpet fibers provide high-frequency elastic resistance.

[0061] Finally, the two-dimensional orthogonal decoupling and fault diagnosis module constructs a two-dimensional orthogonal coordinate system from the amplitude mean and high-frequency variance fluctuation characteristics of the aforementioned pure mechanical jamming torque. Within this coordinate system, the system accurately separates the fault boundaries of "rigid foreign object jamming," "carpet flexible legal fluctuation," and "air duct aerodynamic blockage." Specifically:

[0062] 1. When the average amplitude of the pure mechanical jamming torque is high and the high-frequency variance fluctuation characteristic is extremely small, the system will determine that the floor brush is stuck due to the cable or hair getting caught, and will immediately output an electronic braking command to protect the motor and the user.

[0063] 2. When the average amplitude of the pure mechanical locking torque is high, but the high-frequency variance fluctuation characteristics are large, the system will determine it as high-frequency elastic fluctuation of the floor brush on the long-pile carpet. This is a legal operating condition, and the system will disable the alarm action to avoid false alarms.

[0064] 3. When the average amplitude of the pure mechanical jamming torque approaches zero, but the pneumatic load index is significantly lower than the preset ventilation threshold, the system will determine that the filter or extension tube is completely blocked, and output a frequency reduction operation and filter cleaning prompt command to avoid motor overheating and energy waste.

[0065] In a preferred embodiment of the present invention, the specific formula for the dimensionless standardization process of the core electrical parameter acquisition and dimensionless standardization module is as follows:

[0066] , ,

[0067] in, , , These are the normalized per-unit values ​​of bus voltage, phase current, and motor speed, respectively.

[0068] , , The actual physical absolute value obtained by the sensor or observer;

[0069] , , These are the factory-calibrated reference values ​​for the rated bus voltage, rated phase current, and rated maximum speed of the vacuum cleaner motor.

[0070] Through the dimensionless normalization process, the rotational speed parameters in the maximum numerical domain and electrical parameters of different dimensions are uniformly compressed into a proportional scale, preventing data overflow faults when the microcontroller performs power multiplication and division.

[0071] In this embodiment, the core electrical parameter acquisition and dimensionless per-unit standardization module is the basic data processing unit of the handheld vacuum cleaner stall fault self-diagnosis system. Its core function is to collect key electrical parameters of the motor's operating status in real time and perform standardized processing on them. This module can be implemented by both hardware circuits and software algorithms. The hardware part may include sensors, analog-to-digital converters (ADCs), etc., to convert analog signals into digital signals; the software part is responsible for performing per-unit standardization operations on the collected digital signals. The dimensionless per-unit standardization process converts physical quantities of different dimensions and orders of magnitude into dimensionless per-unit values ​​by dividing the actual measured physical quantity by its corresponding rated reference value.

[0072] Specifically, the per-unit value of bus voltage It is by using the actual bus voltage at the current moment Divide by the rated bus voltage Obtain; per-unit value of phase current It is by using the actual phase current at the current moment Divide by the rated phase current Obtain; per-unit value of motor speed It is by using the actual motor speed at the current moment. Divide by the rated maximum speed reference value These per-unit values ​​are obtained after dimensionless processing, and their numerical range is usually between 0 and 1 (or slightly greater than 1), which greatly simplifies the numerical processing complexity of subsequent algorithms.

[0073] The , , It is the raw, unprocessed physical quantity that is obtained in real time through various detection methods during the actual operation of the brushless DC motor of the handheld vacuum cleaner. It can be measured using a voltage sensor or a voltage divider circuit; It can be measured using a current sensor (such as a Hall sensor or a sampling resistor); This data can be obtained through a motor encoder, a sensorless back EMF observer, or the speed estimation module in a field-oriented control (FOC) algorithm. This raw data serves as input for per-unit processing.

[0074] The , , These are fixed parameters determined during the design and manufacturing of the motor, representing its normal operating range and performance limits. This is typically the DC bus voltage when the motor is operating normally. It is the effective value of the phase current when the motor is running under rated load; This refers to the highest speed that the motor can reach under no-load or light-load conditions.

[0075] Through the above dimensionless standardization process, this application effectively solves the technical problem that in the self-diagnosis system for stall faults of handheld vacuum cleaners, the microcontroller is prone to data overflow when executing complex dynamic equations due to the extremely large numerical range and different dimensions of the motor electrical parameters.

[0076] In a preferred embodiment of the present invention, the specific formula for calculating the aerodynamic load index by the aerodynamic load index non-sensory observation module is as follows:

[0077]

[0078] in, For the current moment The observed aerodynamic load index;

[0079] This represents the duty cycle at the current moment.

[0080] This is the derivative of the per-unit value of the rotational speed, i.e., the per-unit angular acceleration;

[0081] The inverter efficiency coefficient characterizes the power conversion efficiency of the drive circuit. ;

[0082] Per-unit stator phase resistance;

[0083] This is the standardized total moment of inertia of the rotor.

[0084] In this embodiment, the pneumatic load index It is a dimensionless quantity used to quantify the airflow within a vacuum cleaner's duct. Its value directly reflects the duct's unobstructed flow or blockage. This index is a key physical quantity for distinguishing between pneumatic blockages and mechanical jamming. The duty cycle is... This is a crucial parameter in pulse width modulation (PWM) signals, directly controlling the average voltage applied to the brushless DC motor, thus affecting the motor's input power and speed. This parameter is a real-time acquired motor control signal, reflecting the motor's current drive state.

[0085] The standardized angular acceleration This represents the rate of change of the motor's per-unit speed over time, i.e., the degree of acceleration or deceleration of the motor. It is a key parameter for calculating the motor's rotor inertial torque and can be obtained by numerical differentiation of the real-time acquired per-unit speed value or by estimation using a state observer. The inverter efficiency coefficient... This is a constant between 0.85 and 0.95 used to correct for power losses incurred by the inverter during the process of converting DC power into AC power to drive the motor. This coefficient ensures the accuracy of calculating the actual input mechanical power of the motor and can be obtained through experimental calibration or by consulting the inverter's datasheet.

[0086] The per-unit stator phase resistance This is the per-unit value of the motor stator winding resistance, used to calculate the copper losses inside the motor. By subtracting this loss, the power converted into mechanical energy can be obtained more accurately. This value can be obtained from the motor nameplate parameters and then per-unitized, or it can be obtained through motor parameter testing.

[0087] The per-unit total moment of inertia of the rotor This is the standardized value of the total rotational inertia of the motor rotor and centrifugal fan impeller, used to calculate the torque required to overcome inertia during motor acceleration or deceleration. This parameter is crucial for separating the inertial component from the total electromagnetic torque and can be calculated through mechanical design parameters or obtained and standardized through system identification methods.

[0088] This solution achieves precise decoupling and observation of the aerodynamic load index by constructing a mathematical model based on the similarity law of centrifugal fans. In the handheld vacuum cleaner stall fault self-diagnosis system, firstly, the system utilizes the bus voltage per-unit value obtained from the core electrical parameter acquisition and dimensionless per-unit conversion modules. per-unit value of phase current and duty cycle And combined with inverter efficiency coefficient , the actual input electric power of the motor is calculated. This step corrects the energy loss during the power conversion of the drive circuit, ensuring the accuracy of the input power calculation.

[0089] Secondly, the copper loss inside the motor is deducted from the calculated input electric power, that is , where is the per-unit stator phase resistance. This eliminates the interference of motor heating on load observation, making the remaining power more accurately reflect the mechanical load. Then, the system further separates the power component corresponding to the accelerating inertia torque from the remaining power, that is , where is the per-unit total rotor moment of inertia, is the per-unit angular acceleration. Through this compensation, the observation result is no longer affected by the motor acceleration and deceleration process, thus isolating the dynamic inertia effect. Finally, the net mechanical power after the above multiple compensations is divided by the cube of the per-unit rotational speed . This division operation is based on the similarity law that the load torque of the centrifugal fan is proportional to the square of the rotational speed and the power is proportional to the cube of the rotational speed, normalizing the load characteristics into an aerodynamic load index independent of rotational speed fluctuations . This series of processing steps effectively solves the problem of difficult quantification of aerodynamic load under complex working conditions through comprehensive compensation of electrical input, mechanical loss and inertial dynamics, providing a reliable physical quantity basis for subsequent distinguishing between air duct blockage and mechanical jamming.

[0090] As a preferred embodiment of the present invention, the cross-coupling equation for dynamically calculating the per-unit value of the bearing additional friction torque of the axial negative pressure friction force cross-compensation module is specifically:

[0091]

[0092] where, [[ID=2-eight]]is the per-unit value of the bearing additional friction torque including the additional effect of vacuum degree;

[0093] is the per-unit value of the background static friction torque under normal pressure free state;

[0094] is the reference aerodynamic load index of the vacuum cleaner in the state where the air duct is completely unobstructed;

[0095] is the rotor axial force-friction force cross-coupling conversion constant.

[0096] In this embodiment, the axial negative pressure friction cross-compensation module is a key functional unit. Its main responsibility is to dynamically calculate and compensate for the additional bearing friction torque caused by axial negative pressure based on changes in the vacuum cleaner's duct condition. This module can be implemented as part of the microcontroller's internal firmware code by executing specific algorithmic logic.

[0097] The per-unit value of the additional frictional torque of the bearing This is the result of dimensionless processing of the additional frictional torque generated on the motor bearing due to the axial thrust of the fan. This value reflects the dynamic change of the bearing frictional torque under different degrees of duct blockage. Its purpose is to quantify this dynamic frictional interference so that it can be accurately subtracted from the subsequent total electromagnetic torque, thereby obtaining a more realistic mechanical jamming torque.

[0098] The per-unit value of the background static friction torque under normal pressure free state This represents the inherent and stable frictional torque of the vacuum cleaner motor's bearings under ideal, normal-pressure operating conditions with no airflow obstruction and no axial negative pressure. This value is typically obtained through precise experimental measurement and calibration before the vacuum cleaner leaves the factory and is stored as a fixed parameter in the microcontroller's non-volatile memory, providing a basic reference point for dynamic friction compensation.

[0099] The benchmark aerodynamic load index The aerodynamic load index refers to the aerodynamic load index of a vacuum cleaner when the airflow duct is completely unobstructed. It indicates the system's operating condition under optimal airflow conditions. This benchmark value is also determined through standard testing of the vacuum cleaner during the design or production phase and serves as an important reference parameter for the system, used to compare with the real-time observed aerodynamic load index to assess the degree of duct blockage.

[0100] The rotor axial force-friction cross-coupling conversion constant This is a dimensionless proportionality constant that characterizes the quantitative relationship between the change in axial thrust on the rotor and the resulting change in bearing friction torque. This constant is a comprehensive reflection of factors such as the vacuum cleaner's mechanical structure, bearing type, and material properties; its accuracy is crucial for accurately compensating for the additional bearing friction torque. This constant is typically determined through fitting extensive experimental data or physical modeling and is embedded in the system parameters.

[0101] The proposed solution introduces an axial negative pressure friction cross-compensation mechanism to dynamically correct nonlinear friction interference during motor operation. In the above implementation, the core electrical parameter acquisition and dimensionless per-unit conversion module acquires and processes the motor's bus voltage, phase current, duty cycle, and motor speed in real time, converting them into uniform per-unit values. Based on this, the aerodynamic load index sensorless observation module utilizes the centrifugal fan similarity law to remove the inertial component from the electromagnetic torque and calculates an independent aerodynamic load index unaffected by speed fluctuations. This index can represent the actual circulation area of ​​the current air duct.

[0102] Based on this, the axial negative pressure friction cross-compensation module utilizes the aerodynamic load index. The per-unit value of the bearing additional frictional torque is dynamically calculated based on the changes. Specifically, this module, through the established cross-coupling equations, measures the per-unit value of the background static friction torque under normal pressure free state. This serves as the foundation, upon which a dynamic compensation term is superimposed. The core of this dynamic compensation term lies in utilizing the baseline aerodynamic load index. Compared with the real-time observed aerodynamic load index The difference in drop value between them. When the air duct begins to become blocked. It will be lower than Difference The increase reflects the decrease in duct flow area and the enhancement of axial negative pressure. This difference is calculated by squaring the value and multiplying it by the rotor axial force-friction cross-coupling conversion constant. It can simulate the nonlinear characteristics of friction force changing with negative pressure, thereby dynamically calculating the additional bearing friction torque caused by air duct blockage.

[0103] The key to this compensation process lies in decoupling the fluctuations of aerodynamic load from those of mechanical friction. By introducing a cross-coupling conversion constant, the system can dynamically adjust the compensation amount of friction torque in real time according to the actual flow state of the current air duct. In this way, the friction component deducted from the total electromagnetic torque of the motor more closely matches the actual physical working conditions, providing a reliable physical basis for the accurate extraction of pure mechanical jamming torque.

[0104] In a preferred embodiment of the present invention, the specific formula for the pure mechanical jamming torque extraction module to extract the pure mechanical jamming torque is as follows:

[0105]

[0106] in The extracted pure mechanical jamming torque per unit value is a parameter that completely decouples the interference between the duct vacuum and the motor inertia.

[0107] In this embodiment, the inverter efficiency coefficient involved in the scheme This characterizes the energy conversion efficiency of the drive circuit when converting DC power into AC power to drive a brushless DC motor. Bus voltage per unit value. per-unit value of phase current Duty cycle and per-unit value of motor speed All parameters are acquired in real time by the core electrical parameter acquisition and dimensionless per-unit conversion module and then processed into dimensionless per-unit values.

[0108] Per-unit stator phase resistance This represents the resistance value of the motor stator winding, which can be obtained from the motor's factory calibration or through online identification methods. In the calculation of electromagnetic torque, this energy loss caused by resistance needs to be deducted.

[0109] Per-unit rotor total moment of inertia It represents the total inertia of the motor rotor and its connected rotating components such as centrifugal impellers. After being standardized, it is the core parameter for calculating the inertial torque required when the motor accelerates or decelerates. Its value is usually determined by the motor design and can be configured during system initialization.

[0110] Differential of per unit value of rotational speed This represents the rate of change of the per-unit value of motor speed, also known as the per-unit angular acceleration. It reflects the instantaneous trend of motor speed change and is the direct basis for calculating motor inertial torque. This value can be obtained by differential or filtering the real-time per-unit value of motor speed.

[0111] pneumatic load index The aerodynamic load index, calculated by the non-sensory observation module based on the centrifugal fan similarity law, represents the actual flow area of ​​the current air duct, i.e., the magnitude of the aerodynamic resistance of the vacuum cleaner's air duct. Accurate acquisition of this index is crucial for distinguishing between aerodynamic blockage and mechanical jamming.

[0112] Per unit value of additional frictional torque of bearing The aerodynamic-mechanical cross-coupling equation, dynamically calculated by the axial negative pressure friction cross-compensation module based on the rotor axial thrust effect caused by fluid mechanics, reflects the additional frictional torque generated by changes in bearing preload due to negative pressure variations within the air duct (such as air duct blockage). Accurately compensating for this dynamically changing frictional torque is key to improving the accuracy of mechanical jamming torque extraction.

[0113] This application's solution achieves the goal of accurately extracting pure mechanical jamming torque from the total electromagnetic torque of the motor by constructing a comprehensive torque decoupling model. Specifically, it calculates the per-unit value of the motor at the current speed. The total electromagnetic torque. This calculation process first involves... Estimate the effective electrical power input to the motor, and then subtract the power from the input. The stator copper losses are represented to obtain the mechanical power output of the motor, which is then divided by the per-unit speed value. This is converted into total electromagnetic torque. Based on this, in order to obtain pure mechanical locking torque, the system performs a series of precise subtraction operations on the total electromagnetic torque.

[0114] First, subtract the total moment of inertia of the rotor, which has been standardized. The derivative of the per-unit value of the rotational speed The resulting acceleration inertial torque This step effectively eliminates the torque component generated by the inertia effect during motor acceleration or deceleration, ensuring the real-time accuracy of torque extraction and avoiding interference from the motor's dynamic response on jamming judgment.

[0115] Secondly, the system further deducts the aerodynamic load torque. By deducting this, the influence of aerodynamic resistance within the duct on the motor torque is separated from the total torque, thus resolving the interference caused by the fluctuation of aerodynamic load with speed. This ensures that subsequent mechanical jamming judgment is no longer affected by the duct condition.

[0116] Finally, the system also subtracts the per-unit value of the additional bearing friction torque dynamically calculated by the axial negative pressure friction cross-compensation module. This frictional torque is caused by the axial negative pressure generated by the fan impeller during vacuum cleaner operation, which alters the bearing preload and thus affects the frictional torque. This frictional force dynamically changes with the duct blockage state. By precisely compensating for this time-varying frictional component, the interference of mechanical friction fluctuations caused by changes in duct vacuum on the jamming torque is completely eliminated. Through this series of precise subtraction calculations, various non-mechanical jamming interference items (including inertial torque, pneumatic load torque, and dynamic frictional torque affected by vacuum) originally mixed in the total electromagnetic torque are subtracted one by one, finally yielding a pure mechanical jamming torque per unit value. It can accurately reflect the obstructive effect of external physical objects on the motor. This mechanism, which completely decouples the vacuum level of the air duct from the inertial interference of the motor, provides a high signal-to-noise ratio characteristic input for subsequent fault diagnosis, thereby solving the diagnostic bias problem caused by the aerodynamic and mechanical coupling effect in traditional solutions and significantly improving the accuracy and reliability of stall fault diagnosis.

[0117] like Figure 2 As shown, in a preferred embodiment of the present invention, the transient variance fluctuation feature calculation module calculates the high-frequency variance fluctuation features in the following specific way:

[0118] Establish a length of Using a sliding time window, calculate the high-frequency variance fluctuation characteristics of the pure mechanical jamming torque within that window. :

[0119]

[0120] in This represents the average amplitude of the per-unit torque value within the sliding time window.

[0121] In this embodiment, the transient variance fluctuation feature calculation module is a functional unit specifically designed to process time-series data and extract its statistical features. Its main function is to quantify the intensity of signal fluctuations, thereby providing a key criterion for distinguishing different types of faults. This module can be implemented in various ways; for example, it can be a piece of firmware code in a microcontroller, processing the acquired data through software algorithms; or it can be designed as a dedicated digital signal processor or field-programmable gate array module to achieve higher speed and parallel computing capabilities. The sliding time window of length N is a data processing mechanism designed to perform local and dynamic analysis on continuously input data streams. The introduction of this window ensures the real-time nature of fault diagnosis and the ability to capture transient changes. In specific implementation, the sliding time window can be a first-in, first-out queue maintained in system memory to store the most recent N data points; or it can be a circular buffer, managing data entry and exit through a fixed-size array and pointers to optimize memory usage and data access efficiency. The high-frequency variance fluctuation feature of the pure mechanical jamming torque... This is a statistic that precisely characterizes the dispersion or fluctuation severity of pure mechanical jamming torque within a sliding window. As a key indicator distinguishing between rigid jamming and flexible fluctuation faults, its calculation can be strictly performed according to a given mathematical formula: summing the squares of the differences between each data point within the window and the mean, and then averaging the sums. To improve computational efficiency, a recursive or incremental algorithm can also be used to avoid recalculating the mean and variance of the entire window each time. The mean amplitude of the per-unit torque value within the sliding time window is... This represents the average value of the pure mechanical jamming torque within the window. It serves not only as a benchmark for calculating variance but also as another important dimension for fault diagnosis. It can be calculated by summing all data points within the sliding window and then dividing by N; alternatively, it can be updated using the mean from the previous moment and the new and old data points through a recursive mean method to reduce computational load.

[0122] This application's solution introduces transient variance fluctuation characteristics, providing a time-domain statistical criterion for fault diagnosis, thus effectively solving the problem of fuzzy diagnosis under complex operating conditions caused by a single amplitude index. The solution receives pure mechanical jamming torque output from the pure mechanical jamming torque extraction module, which has been completely decoupled from the duct vacuum level and motor inertial interference. This allows subsequent fluctuation characteristic analysis to focus more on the characteristics of the mechanical jamming itself, avoiding interference from other factors. By establishing a sliding time window of length N, the system can continuously and dynamically monitor the pure mechanical jamming torque. This windowing processing method ensures that the diagnostic logic can respond to instantaneous changes in torque in real time, rather than relying solely on static average values. Calculating the discrete high-frequency variance fluctuation characteristics of the torque within this window has a core contribution in quantifying the intensity of torque signal fluctuations. Since rigid foreign object jamming typically manifests as a sustained high amplitude torque with relatively small fluctuations, while elastic fluctuations caused by carpet fibers manifest as high-frequency and severe torque oscillations, calculating the variance transforms this difference in physical characteristics into quantifiable mathematical features. By using the mean amplitude within a sliding window as a benchmark, the square and average values ​​of the torque deviations are calculated. This process precisely isolates the DC component of the torque, allowing the diagnostic system to focus on extracting high-frequency disturbance information from the torque signal. This provides crucial feature input for the subsequent precise decoupling of rigid jamming and flexible ripple. This method, combining pure torque input and time-domain ripple feature analysis, significantly improves the accuracy and robustness of fault diagnosis.

[0123] In a preferred embodiment of the present invention, the logic for performing two-dimensional orthogonal decoupling and fault diagnosis by the two-dimensional orthogonal decoupling and fault diagnosis module specifically includes:

[0124] (1) Rigid mechanical jamming: when Exceeding the safety warning threshold and exhibiting high-frequency variance fluctuation characteristics If the system determines that the floor brush is stuck due to cable or hair getting caught in it, it will output an electronic braking command.

[0125] (2) False alarm filtering for carpet flexibility: When Exceeding the safety warning threshold and exhibiting high-frequency variance fluctuation characteristics When the system detects high-frequency elastic fluctuations of the floor brush on the long-pile carpet, it determines the condition as a valid working condition and disables the alarm.

[0126] (3) Air duct aerodynamic blockage: When the average value of pure mechanical jamming torque is Approaching zero, and the aerodynamic load index When the ventilation threshold is lower than the preset threshold, it is determined that the filter or extension tube is completely blocked, and a frequency reduction and filter cleaning prompt command is output.

[0127] in The smallest first variance constant characterizing the rigid drag properties, Let be the pre-defined second variance constant characterizing the elastic fluctuation resistance properties of long-pile carpets, and , ; .

[0128] In this embodiment, the average amplitude of the pure mechanical locking torque This represents the average magnitude of the pure mechanical jamming torque caused by external physical foreign objects, after deducting the per-unit values ​​of acceleration inertia torque, aerodynamic load torque, and bearing-related frictional torque. Its function is to provide a stable indication of the level of mechanical resistance, unaffected by other interfering factors. This average value can be obtained by calculating the arithmetic mean of the pure mechanical jamming torque within a sliding time window, or by obtaining a smoothed value through low-pass filtering or other methods.

[0129] High-frequency variance fluctuation characteristics The discrete high-frequency fluctuations of the pure mechanical jamming torque in the time domain are quantified. It reflects the drastic change and frequency characteristics of mechanical resistance, and is a key indicator distinguishing between rigid jamming and flexible fluctuations. This characteristic can be obtained by calculating the variance of the pure mechanical jamming torque within a sliding time window, or by extracting the high-frequency components of the torque signal and calculating its energy or root mean square value. A safety warning threshold is used to determine whether the average amplitude of the pure mechanical jamming torque reaches a dangerous level requiring attention. When the average torque exceeds this threshold, it indicates a potential abnormality in mechanical resistance. This threshold can be set according to the rated load, maximum permissible overload capacity, and safety margin of the vacuum cleaner motor; for example, it can be set as a percentage of the motor's rated torque, or determined through experimental testing under different operating conditions.

[0130] First variance constant This is an extremely small value used to characterize rigid resistance characteristics. When the high-frequency variance fluctuation characteristics are below this constant, combined with the high torque mean, it can be determined as rigid jamming. This constant can be determined by analyzing the actual operating data of the vacuum cleaner under rigid foreign object jamming conditions and statistically analyzing the upper limit of its torque fluctuation characteristics, or by optimizing it through a simulation model.

[0131] Second variance constant This constant is used to characterize the elastic fluctuation resistance properties of long-pile carpets. When the high-frequency variance fluctuation characteristics exceed this constant, combined with the high torque mean, it can be determined as carpet flexibility fluctuation. This constant can be determined by collecting and analyzing actual data of vacuum cleaners running on long-pile carpets to statistically determine the lower limit of their torque fluctuation characteristics, or it can be set through expert experience.

[0132] pneumatic load index This index characterizes the actual flow area of ​​the current air duct, i.e., the load status of the vacuum cleaner's pneumatic system. Its function is to distinguish between mechanical and pneumatic blockages. This index can be calculated using an aerodynamic observer based on the centrifugal fan similarity law, for example, by measuring parameters such as the pressure difference between the fan inlet and outlet, and flow rate, or, as described above, by removing the inertial component from the per-unit value of the electromagnetic torque and dividing it by the cube of the per-unit value of the rotational speed. The ventilation threshold is used to determine whether the pneumatic load index is below the minimum level under normal ventilation conditions, thus indicating whether the air duct is severely blocked. When the pneumatic load index is below this threshold, it indicates that the air duct flow area has significantly decreased. This threshold can be experimentally calibrated based on the pneumatic load index under different air duct conditions such as normal operation, slight blockage, and complete blockage, or theoretically calculated based on the fan performance curve. The electronic brake command is used to quickly stop the motor when rigid jamming is detected, to protect the motor and user safety. This command can be implemented by sending a stop signal to the inverter drive circuit or by rapidly reducing the motor's supply voltage or current. The frequency reduction and filter cleaning reminder command is used to reduce motor speed to decrease energy consumption and reduce the risk of motor overheating when aerodynamic blockage of the air duct is detected, and to remind the user to clean the filter. This command can achieve frequency reduction by sending a signal to the inverter drive circuit to reduce the duty cycle, and at the same time, it will give the user a cleaning reminder through the display screen, indicator lights or buzzer.

[0133] The solution proposed in this application constructs a two-dimensional orthogonal decoupling and fault diagnosis module to average the amplitude of the pure mechanical jamming torque. High-frequency variance fluctuation characteristics As two orthogonal dimensions, combined with the aerodynamic load index This system enables accurate classification and diagnosis of complex operating conditions of vacuum cleaners. Specifically, the system first acquires the motor's bus voltage, phase current, duty cycle, and motor speed in real time through the core electrical parameter acquisition and dimensionless per-unit conversion module, and then performs per-unit conversion to provide input with a unified dimension for subsequent calculations. Next, the aerodynamic load index sensorless observation module calculates the independent aerodynamic load index, which is unaffected by speed fluctuations, based on the centrifugal fan similarity law. This is used to characterize the actual flow area of ​​the current air duct. Simultaneously, the axial negative pressure friction cross-compensation module dynamically calculates the per-unit value of the bearing additional friction torque caused by air duct blockage based on the rotor axial thrust effect caused by fluid mechanics. On this basis, the pure mechanical jamming torque extraction module accurately subtracts the acceleration inertia torque, aerodynamic load torque, and the aforementioned bearing additional friction torque per-unit value from the total electromagnetic torque per-unit value of the motor, thereby obtaining the pure mechanical jamming torque that completely decouples the air duct vacuum degree from the motor inertial interference. Subsequently, the transient variance fluctuation characteristic calculation module establishes a sliding time window to calculate the discrete high-frequency variance fluctuation characteristics of the pure mechanical jamming torque in the time domain. and its mean amplitude Finally, the two-dimensional orthogonal decoupling and fault diagnosis module averages the amplitude of the pure mechanical jamming torque. With the aforementioned high-frequency variance fluctuation characteristics Construct a two-dimensional orthogonal coordinate system and combine it with the aerodynamic load index. Perform logical judgment. When Greater than the safety warning threshold, and Less than the smallest first variance constant If the system determines that the floor brush is stuck due to cable or hair getting caught, it will immediately output an electronic brake command to prevent damage to the motor and injury to the user. Greater than the safety warning threshold, but Greater than the preset second variance constant At that time, the system recognized that this was a high-frequency elastic fluctuation of the floor brush on a long-pile carpet, which is a legal operating condition. Therefore, it disabled alarm activation, preventing accidental shutdowns and improving the user experience. Furthermore, when the average amplitude of the pure mechanical jamming torque... Approaching zero, and the aerodynamic load index When the ventilation threshold is lower than preset, the system determines that the filter or extension tube is completely blocked, and outputs a frequency reduction and filter cleaning prompt, effectively solving the problem that traditional solutions cannot identify "underload" type duct blockage faults. This multi-dimensional and refined diagnostic logic enables the system to accurately distinguish different types of faults and take targeted intervention measures.

[0134] like Figure 3 As shown, a handheld vacuum cleaner stall fault self-diagnosis device applies the aforementioned handheld vacuum cleaner stall fault self-diagnosis system, including a vacuum cleaner body, and a brushless DC motor, a centrifugal impeller in the air duct, an inverter drive circuit, and a sensorless sampling circuit installed inside the vacuum cleaner body; a microcontroller, in which the handheld vacuum cleaner stall fault self-diagnosis system is burned in firmware code; the microcontroller reads the electrical signals of the inverter drive circuit through the sensorless sampling circuit, uses the built-in stall fault self-diagnosis system to output the current fault classification command in real time, and drives the inverter drive circuit to perform physical protection actions such as electronic brake stop, low duty cycle reverse, or frequency reduction operation.

[0135] The core innovation of this embodiment lies in combining the amplitude mean of the pure mechanical jamming torque with the transient variance fluctuation characteristics in a two-dimensional orthogonal coordinate system, and introducing an axial negative pressure friction cross-compensation mechanism to accurately separate the fault boundaries of rigid jamming, carpet flexible fluctuation, and air duct blockage. The system first acquires electrical parameters such as bus voltage, phase current, duty cycle, and motor speed through a sensorless sampling circuit. After dimensionless normalization, an independent parameter characterizing the air duct flow area is extracted from the electromagnetic torque using a pneumatic load index sensorless observation module. Simultaneously, the axial negative pressure friction cross-compensation module dynamically calculates and compensates for the additional bearing friction torque caused by air duct blockage based on the drop difference of the pneumatic load index. On this basis, the pure mechanical jamming torque extraction module subtracts the acceleration inertia torque, pneumatic load torque, and compensated friction torque from the total electromagnetic torque to obtain a pure component reflecting only the resistance of external physical foreign objects. The transient variance fluctuation characteristic calculation module then analyzes the time-domain fluctuation characteristics of this torque through a sliding time window. Finally, the two-dimensional orthogonal decoupling and fault diagnosis module maps the amplitude mean and high-frequency variance features to an orthogonal coordinate system to achieve accurate classification of fault types: high amplitude mean and low variance correspond to rigid jamming triggering electronic braking; high amplitude mean and high variance correspond to carpet fluctuation maintaining operation; low amplitude mean combined with a decrease in aerodynamic load index corresponds to duct blockage triggering frequency reduction prompt.

[0136] The above technical solutions effectively overcome the limitations of traditional methods that rely on a single threshold, avoid frequent accidental shutdowns during carpet cleaning, ensure timely protection in case of rigid jamming, and identify underload faults caused by air duct blockage, significantly improving the intelligence level of the vacuum cleaner and the user experience.

[0137] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A self-diagnostic system for stall faults in a handheld vacuum cleaner, characterized in that, include: Core electrical parameter acquisition and dimensionless per-unit conversion module: used to acquire the bus voltage, phase current, duty cycle and motor speed of the brushless DC motor of the handheld vacuum cleaner in real time, and to perform dimensionless per-unit conversion on the above physical quantities using the rated reference parameters of the motor, mapping them to per-unit values. Aerodynamic load index non-intrusive observation module: used to establish an aerodynamic observer based on the centrifugal fan similarity law, remove the inertial component from the per-unit value of electromagnetic torque, divide it by the cube of the per-unit value of speed, and calculate the independent aerodynamic load index that is not affected by speed fluctuations, so as to characterize the true flow area of ​​the current air duct. Axial negative pressure friction cross compensation module: used to establish aerodynamic-mechanical cross coupling equation based on the rotor axial thrust effect caused by fluid mechanics; and dynamically calculate the per-unit value of the bearing additional friction torque caused by air duct blockage using the drop difference value of the aerodynamic load index. Pure mechanical jamming torque extraction module: used to accurately deduct the acceleration inertia torque, aerodynamic load torque and the bearing additional friction torque per unit value from the total electromagnetic torque per unit value of the motor, and extract the pure mechanical jamming torque caused only by the entrainment of external physical foreign objects; Transient variance fluctuation characteristic calculation module: used to establish a sliding time window and calculate the discrete high-frequency variance fluctuation characteristics of the pure mechanical jamming torque in the time domain; Two-dimensional orthogonal decoupling and fault diagnosis module: used to construct a two-dimensional orthogonal coordinate system by combining the amplitude mean of the pure mechanical jamming torque with the high-frequency variance fluctuation characteristics, accurately separate the fault boundaries of "rigid foreign object jamming", "carpet flexible legal fluctuation" and "air duct aerodynamic blockage", and output classification intervention instructions.

2. The system according to claim 1, characterized in that, The specific method for the dimensionless per-unit processing of the core electrical parameter acquisition and dimensionless per-unit conversion module is as follows: divide the real-time acquired bus voltage, phase current and motor speed by the corresponding rated bus voltage, rated phase current and rated maximum speed to obtain their respective per-unit values.

3. The system according to claim 2, characterized in that, The specific method by which the aerodynamic load index non-sensory observation module calculates the aerodynamic load index is as follows: First, calculate the inverter efficiency coefficient by multiplying the bus voltage per unit value, the phase current per unit value, and the duty cycle by the product. Then, subtract the square of the phase current per unit value multiplied by the per-unit stator phase resistance, and then subtract the per-unit rotor total moment of inertia multiplied by the derivative of the speed per unit value with respect to time, and then multiply by the speed per unit value. Then divide the above result by the cube of the per-unit speed value to obtain the aerodynamic load index at the current moment.

4. The system according to claim 3, characterized in that, The cross-coupling equation for the dynamic calculation of the per-unit value of the bearing's additional frictional torque by the axial negative pressure friction cross-compensation module is as follows: The per-unit value of the base static friction torque under normal pressure free state is added to the square of the rotor axial force-friction force cross-coupling conversion constant multiplied by the positive part of the difference between the reference aerodynamic load index and the current aerodynamic load index, to obtain the per-unit value of the bearing additional friction torque including the vacuum degree additional effect.

5. The system according to claim 4, characterized in that, The specific formula for extracting pure mechanical jamming torque by the pure mechanical jamming torque extraction module is as follows: First, calculate the per-unit value of the total electromagnetic torque of the motor. This value is equal to the inverter efficiency coefficient multiplied by the per-unit value of the bus voltage, the per-unit value of the phase current, and the duty cycle in sequence. Subtract the square of the per-unit value of the phase current multiplied by the per-unit stator phase resistance from the result, and then divide the difference by the per-unit value of the speed. Then, subtract the acceleration inertia torque, aerodynamic load torque, and bearing additional friction torque per unit value from the total electromagnetic torque per unit value in sequence. The acceleration inertia torque is equal to the per-unit rotor total moment of inertia multiplied by the differential of the per-unit speed value, and the aerodynamic load torque is equal to the aerodynamic load index multiplied by the square of the per-unit speed value, finally obtaining the pure mechanical jamming torque per unit value.

6. The system according to claim 5, characterized in that, The transient variance fluctuation feature calculation module calculates the high-frequency variance fluctuation features in the following specific way: Establish a sliding time window of length N, calculate the mean amplitude of the pure mechanical jamming torque within the window; then calculate the square of the difference between the torque value and the mean at all times within the window, sum these squared values, and divide by the window length N to obtain the high-frequency variance fluctuation characteristics.

7. The system according to claim 6, characterized in that, The logic for performing two-dimensional orthogonal decoupling and fault diagnosis by the two-dimensional orthogonal decoupling and fault diagnosis module specifically includes: When the average amplitude of the pure mechanical jamming torque is greater than the safety warning threshold and the high-frequency variance fluctuation characteristic is less than the first variance constant, it is determined to be rigid mechanical jamming, and an electronic braking command is output. When the average amplitude is greater than the safety warning threshold and the high-frequency variance fluctuation characteristic is greater than the second variance constant, it is judged as a carpet flexible legal fluctuation and the alarm action is blocked. When the average amplitude approaches zero and the aerodynamic load index is lower than the preset ventilation threshold, it is determined that the air duct is aerodynamically blocked, and a frequency reduction and filter cleaning prompt command is output. The first variance constant is smaller than the second variance constant.

8. A self-diagnostic device for stall faults in a handheld vacuum cleaner, characterized in that, The system described in any one of claims 1 to 7 comprises: The vacuum cleaner body, and the brushless DC motor, air duct centrifugal impeller, inverter drive circuit and sensorless sampling circuit installed inside the vacuum cleaner body; The handheld vacuum cleaner stall fault self-diagnosis system is burned into the microcontroller in the form of firmware code; The microcontroller reads the electrical signal of the inverter drive circuit through the sensorless sampling circuit, uses the built-in stall fault self-diagnosis system to output the current fault classification command in real time, and drives the inverter drive circuit to perform physical protection actions such as electronic brake stop, low duty cycle reverse or frequency reduction operation.

Images