Dust collector control method, control system and dust collector

Abstract

The application discloses a dust collector control method, a control system and a dust collector, and belongs to the technical field of intelligent cleaning equipment. The dust collector control method comprises the following steps: synchronously collecting vibration signals of the air duct system and a holding part; performing time-frequency domain joint analysis on the vibration signals, and extracting a plurality of characteristic values respectively representing airflow state in the air duct, contact state of the floor brush and the ground; performing multi-modal data fusion on the characteristic values and real-time working current of the motor, inputting the multi-modal data fusion into a preset recognition model for analysis, accurately identifying the current working state of the dust collector, including but not limited to the type of air duct blockage, the lifting state of the floor brush and the change of the ground material; and dynamically adjusting the working parameters of the motor according to the identified specific working state. The application can realize high-precision identification of different blockage types and external interference of the dust collector, and is suitable for different ground materials and complex working condition changes, so that better cleaning efficiency and user experience are achieved under a low-cost sensor architecture.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent cleaning equipment, in particular to a dust collector control method, a control system and a dust collector. BACKGROUND

[0002] In the running process of the existing dust collector, the system working state is usually indirectly judged by relying on motor current change, air pressure or flow sensor in the air duct, and the suction force is adjusted or fault prompt is given accordingly. However, this kind of method has obvious limitations: on the one hand, the air pressure or flow sensor has high cost, and is easily disturbed and has insufficient stability under complex working conditions; on the other hand, it is difficult to accurately distinguish different fault types such as air inlet blockage, air outlet blockage or filter screen blockage only by a single physical quantity, and it is also impossible to effectively identify the low-frequency mechanical disturbance caused by the user dragging and swinging the dust collector handle, so that the system mistakenly identifies the human operation as a fault, frequently triggers unnecessary protection or adjustment, and seriously affects the user experience and the reliability of the automatic function. SUMMARY

[0003] The main purpose of the present application is to provide a dust collector control method, a control system and a dust collector, which solve the technical problems that the existing dust collector cannot accurately distinguish fault types and is easily mistaken under user operation interference.

[0004] To achieve the above-mentioned purpose, the present application provides a dust collector control method applied to a dust collector, wherein the dust collector comprises an air duct system, a holding part, a motor and a plurality of sensors, the plurality of sensors are distributed in at least two key parts of the air duct system and the holding part, and the dust collector control method comprises the following steps:

[0005] The vibration signals of at least two key parts in the air duct system and the vibration signal of the holding part are synchronously collected by the plurality of sensors;

[0006] The collected vibration signals are processed to extract feature values representing the airflow state inside the air duct system and feature values representing the contact state between the brush of the dust collector and the ground;

[0007] The extracted feature values are fused with the real-time working current of the motor and input into a preset identification model for analysis to identify the current working state of the dust collector;

[0008] The working parameters of the motor are adjusted according to the identified current working state.

[0009] In an embodiment, the air duct system comprises an air inlet, an air duct middle part and a filter rear part arranged along the air flow propagation direction; wherein the sensor comprises a plurality of piezoelectric sensors and a low-frequency vibration sensor, the plurality of piezoelectric sensors are arranged at the air inlet, the air duct middle part and the filter rear part to collect vibration signals of the corresponding parts; the low-frequency vibration sensor is arranged at the holding part to collect low-frequency vibration signals of the holding part; the specific steps of synchronously collecting vibration signals of at least two key parts in the air duct system and vibration signals of the holding part through the plurality of sensors comprise:

[0010] The vibration signals of the air inlet, the air duct middle part and the filter rear part are synchronously collected through the plurality of piezoelectric sensors, and are numbered as P1 signal, P2 signal and P3 signal respectively;

[0011] The low-frequency vibration signals of the holding part are collected through the low-frequency vibration sensor, and are numbered as P4 signal.

[0012] In an embodiment, the specific steps of performing signal processing on the collected vibration signals to extract characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the vacuum cleaner with the ground comprise:

[0013] The collected vibration signals are preprocessed to obtain preprocessed vibration signals;

[0014] The preprocessed vibration signals are subjected to time-frequency transformation to obtain corresponding frequency domain signals;

[0015] Characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the vacuum cleaner with the ground are extracted from the frequency domain signals.

[0016] In an embodiment, the specific steps of extracting characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the vacuum cleaner with the ground from the frequency domain signals comprise:

[0017] A first characteristic value is extracted from the frequency domain signal of the P1 signal, and the first characteristic value is used to represent the airflow state of the air inlet;

[0018] A second characteristic value is extracted from the frequency domain signals of the P2 signal and the P3 signal, and the second characteristic value is used to represent the balance state of the air duct pressure;

[0019] A third characteristic value is extracted from the frequency domain signal of the P4 signal, and the third characteristic value is used to represent the contact state of the brush of the vacuum cleaner with the ground.

[0020] In an embodiment, the specific step of extracting a second characteristic value from the frequency domain signals of the P2 signal and the P3 signal for characterizing the balance state of the air duct pressure comprises:

[0021] extracting first amplitude data in a first predetermined frequency point or a first narrow frequency band from the frequency domain signal of the P2 signal;

[0022] extracting second amplitude data corresponding to the first predetermined frequency point or the first narrow frequency band from the frequency domain signal of the P3 signal;

[0023] calculating the ratio of the first amplitude data and the second amplitude data to obtain a second characteristic value for characterizing the balance state of the air duct pressure.

[0024] In an embodiment, the specific step of fusing the extracted characteristic value with the real-time working current of the motor and inputting into a preset recognition model for analysis to identify the current working state of the vacuum cleaner comprises:

[0025] obtaining the working current change condition of the motor;

[0026] if the third characteristic value is less than a first preset threshold value, and the first characteristic value is less than a second preset threshold value, it is determined that the air inlet of the vacuum cleaner is blocked;

[0027] if the second characteristic value is greater than a third preset threshold value, and the working current of the motor continues to slowly rise, it is determined that the filter screen of the vacuum cleaner is blocked;

[0028] if the third characteristic value is not less than the first preset threshold value, the first characteristic value is not less than the second preset threshold value, and the working current of the motor sharply rises and exceeds a preset rated value, and the frequency domain signals of the P2 signal and the P3 signal are abnormal, it is determined that the exhaust of the vacuum cleaner is blocked.

[0029] In an embodiment, the specific step of adjusting the working parameters of the motor according to the identified current working state comprises:

[0030] when it is determined that the air inlet of the vacuum cleaner is blocked, the motor is controlled to perform an intermittent super-power operation mode, and the user is prompted;

[0031] when it is determined that the filter screen of the vacuum cleaner is blocked, the power of the motor is reduced, and the user is prompted;

[0032] when it is determined that the exhaust of the vacuum cleaner is blocked, the power of the motor is reduced to a safety gear, and an audible and visual alarm is issued and the user is prompted.

[0033] In an embodiment, the dust collector comprises a pneumatic adjusting mechanism arranged in the air duct system; the dust collector control method further comprises:

[0034] According to the identified working state, the pneumatic adjusting mechanism is controlled to act to adjust the air flow state of the air duct system.

[0035] In addition, to achieve the above-mentioned purpose, the present application further proposes a control system, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the dust collector control method as described above.

[0036] In addition, to achieve the above-mentioned purpose, the present application further proposes a dust collector, comprising the control system as described above, and

[0037] an air duct system;

[0038] a holding portion;

[0039] a motor;

[0040] a plurality of sensors arranged respectively at different positions in the air duct system and on the holding portion, for collecting vibration signals of different positions in the air duct system and vibration signals of the holding portion;

[0041] a motor driving module, an input end of which is electrically connected to an output end of the control system, and an output end of which is electrically connected to the motor, for receiving control signals of the control system to adjust the power of the motor;

[0042] a human-computer interaction module, which is electrically connected to the control system, for displaying the working state of the dust collector.

[0043] The one or more technical solutions proposed by the present application have at least the following technical effects:

[0044] The dust collector control method provided in the application synchronously collects vibration signals through sensors arranged at multiple key positions of an air duct system and a holding part, and combines motor currents for multi-source information fusion, to realize high-precision identification of different blockage types and external interference of the dust collector. Specifically, joint signal processing and feature extraction are performed on multiple vibration signals, which can effectively separate specific vibration modes caused by different faults such as air inlet blockage, air duct blockage or filter saturation, and simultaneously effectively capture and distinguish low-frequency disturbance signals introduced by user operation through the holding part sensor, to reduce the misjudgment rate; further, a preset identification model is used to intelligently analyze multi-dimensional features and current information, to accurately determine the actual working state of the dust collector, and then realize closed-loop optimization and adjustment of motor parameters. This method not only greatly improves the reliability of fault identification and anti-interference, avoids invalid adjustment and energy waste, but also can adapt to different ground materials and working conditions, so as to achieve better cleaning efficiency and user experience under a low-cost sensor architecture. BRIEF DESCRIPTION OF DRAWINGS

[0045] The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate embodiments consistent with the present application and, together with the description, further serve to explain the principles of the application.

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor.

[0047] Figure 1 A flowchart provided for the dust collector control method embodiment one of the present application;

[0048] Figure 2 A flowchart provided for the dust collector control method embodiment two of the present application;

[0049] Figure 3 A flowchart provided for the dust collector control method step S200 embodiment three of the present application;

[0050] Figure 4 A flowchart provided for the dust collector control method embodiment four of the present application;

[0051] Figure 5 A schematic diagram of the composition structure of the dust collector in the embodiments of the present application.

[0052] Explanation of reference numerals: main control MCU 01, motor driving module 02, motor 03, man-machine interaction module 04, pneumatic adjustment mechanism 05, air duct system 06, sensor array 07.

[0053] The objectives, functional features and advantages of the present application will be further illustrated in conjunction with the embodiments, with reference to the accompanying drawings. DETAILED DESCRIPTION

[0054] It should be understood that the specific embodiments described herein are merely intended to explain the technical solutions of the present application, and are not intended to limit the present application.

[0055] In order to better understand the technical solutions of the present application, the specific embodiments will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0056] The present application proposes a dust collector control method, as shown in Figure 1 The dust collector control method comprises the following steps:

[0057] S100: synchronously collecting vibration signals of at least two key positions in the air duct system 06 and vibration signals of the holding part through the plurality of sensors;

[0058] S200: performing signal processing on the collected vibration signals to extract characteristic values representing the airflow state inside the air duct system 06 and characteristic values representing the contact state between the brush of the dust collector and the ground;

[0059] S300: fusing the extracted characteristic values with the working current of the motor 03 and inputting them into a preset recognition model for analysis to identify the current working state of the dust collector;

[0060] S400: adjusting the working state of the motor 03 according to the identified current working state.

[0061] More specifically, the dust collector mainly consists of a power system and an air duct system 06, and the core of the power system is a motor 03 and a fan. The motor 03 provides rotary power to drive the fan to rotate at high speed, thereby generating strong negative pressure in the closed air duct system 06, which is the fundamental principle of the dust collector. The air duct system 06 is a closed path for air flow inside the dust collector, including an air inlet, a dust collection bin, a filter screen, a fan chamber and an air outlet. The air inlet, as the starting point of the air duct, is usually located in the brush or cleaning head, responsible for sucking dirty air and garbage; the dust collection bin serves as an air transfer station, causing a sudden drop in flow rate, and realizing dust separation and collection by centrifugal force or gravity; the filter screen is located behind the dust collection bin and in front of the motor 03, used for filtering small particles, and its blockage will cause significant changes in wind resistance and vibration; the fan chamber contains the motor 03 and the fan, which is the core area of generating suction force; the air outlet is the discharge channel for clean air.

[0062] The existing dust collector usually relies on the current change of the motor 03, the air pressure or flow sensor in the air duct to indirectly judge the system working state during operation, and adjusts the suction force or gives a fault prompt accordingly. However, such a method has obvious limitations: on the one hand, the air pressure or flow sensor is relatively high in cost, and is easily disturbed and insufficient in stability under complex working conditions; on the other hand, it is difficult to accurately distinguish different fault types such as air inlet blockage, air outlet blockage or filter screen blockage only by a single physical quantity, and it is also impossible to effectively identify the low-frequency mechanical disturbance caused by the user dragging and swinging the handle of the dust collector, so that the system mistakenly identifies the human operation as a fault, frequently triggers unnecessary protection or adjustment, and seriously affects the user experience and the reliability of the automatic function. Therefore, how to accurately distinguish various blockage conditions under low-cost conditions and effectively eliminate user operation interference has become a key problem to improve the intelligent control level of the dust collector.

[0063] The reason is that the existing dust collector relies on traditional physical sensors in the state perception aspect, which is limited in information dimension and difficult to capture the subtle differences of different blockage types in vibration frequency spectrum and time domain characteristics; at the same time, due to the lack of effective monitoring of user operation behavior, the system easily confuses the low-frequency vibration of the handle caused by human movement with the high-frequency vibration caused by real blockage, resulting in misjudgment. The essence of this problem lies in the failure to realize the synchronous acquisition and fusion analysis of multi-source and multi-position signals, and the lack of feature separation mechanism for operation interference.

[0064] To break through this technical difficulty, the present application proposes to use low-cost piezoelectric sensors, which are respectively deployed at the air inlet, the middle part of the air duct, the rear of the filter screen and the handle of the dust collector, to obtain vibration signals through multi-point synchronous sampling, and introduce a blind source separation algorithm to effectively extract the characteristic components corresponding to different blockage sources from the mixed signals, so as to realize low-cost and high-reliability blockage type discrimination. At the same time, the detection and analysis of the low-frequency vibration characteristics of the handle are particularly added, so as to distinguish the interference signals introduced by human operation from the real fault response, improve the anti-interference ability and state recognition accuracy of the system in actual use scenarios. This method not only overcomes the shortcomings of high cost and misjudgment of traditional solutions, but also provides a more stable and reliable technical path for realizing intelligent adjustment of the dust collector.

[0065] The following are the specific steps of the present embodiment:

[0066] Step S100 synchronously collects vibration signals of at least two different parts in the air duct system 06 and the vibration signal of the holding part through multiple sensors. In this step, the sensors arranged at key nodes such as the air inlet and the air duct are used to capture the vibration response caused by airflow changes, blockage or filter load inside the air duct system 06; at the same time, the holding part sensor is responsible for monitoring the mechanical vibration generated in user operation, such as low-frequency interference caused by dragging, collision or posture change. Through high-precision synchronous sampling, the system obtains multiple channels of original vibration data with time alignment characteristics, providing a basis for subsequent signal analysis and feature extraction.

[0067] In step S200, the collected original vibration signals are preprocessed and features are extracted. The preprocessing includes filtering, denoising and blind source separation algorithm to eliminate the influence of environmental noise and irrelevant mechanical vibration, and separate the pure "source signal". Feature extraction is designed for two types of key states: one is to reflect the airflow dynamics in the air duct, such as energy distribution in a specific frequency band, etc., which is used to infer whether blockage or airflow restriction occurs; the other is to characterize the contact state of the brush with the ground, such as time-domain impact response, low-frequency vibration intensity, etc., which is used to determine whether the vacuum cleaner is on a carpet, a hard surface or a complex moving state. These features together constitute the information basis for identifying the working state of the system.

[0068] In the S300 stage, the system combines the vibration features with the real-time working current of the motor 03 and inputs them into a preset identification model, such as a classifier based on machine learning or a rule-based decision logic, for comprehensive analysis. The motor 03 current reflects the load change of the whole machine, which can be checked with the vibration features; for example, a current rise accompanied by the disappearance of a specific vibration mode of the air duct may indicate an air inlet blockage, while a slight current fluctuation combined with a large amplitude low-frequency vibration of the handle may be judged as a result of human dragging. The model accurately identifies the current working state category, such as normal cleaning, inlet blockage, exhaust pipe blockage or abnormal user operation, through multi-signal fusion and pattern matching.

[0069] Finally, in the S400 stage, the control system dynamically adjusts the output power or operating mode of the motor 03 according to the identified working state category. For example, when the air inlet blockage is identified, the current is automatically reduced and a warning is triggered; when the brush is lifted, the suction is reduced to save energy; and when human shaking is determined, the current power is maintained to avoid misadjustment. Through this closed-loop adjustment mechanism, the cleaning performance is guaranteed while the adaptability to complex use scenarios and the overall reliability of the system are improved.

[0070] The dust collector control method provided in the application synchronously collects vibration signals through sensors arranged at multiple key positions of the air duct system 06 and the holding portion, and combines motor 03 current for multi-source information fusion to realize high-precision identification of different blockage types of the dust collector and external interference. Specifically, joint signal processing and feature extraction are performed on multiple vibration signals to effectively separate specific vibration modes caused by different faults such as air inlet blockage, air duct blockage, or filter saturation, and at the same time, the holding portion sensor effectively captures and distinguishes low-frequency disturbance signals introduced by user operation, thereby reducing the misjudgment rate; further, the preset identification model is used to intelligently analyze multi-dimensional features and current information to accurately determine the actual working state of the dust collector, and then realize closed-loop optimization and adjustment of the motor 03 parameters. This method not only greatly improves the reliability of fault identification and anti-interference, avoids invalid adjustment and energy waste, but also adapts to different ground materials and working conditions, thereby achieving better cleaning efficiency and user experience under a low-cost sensor architecture.

[0071] In an embodiment, the air duct system 06 includes an air inlet, a middle air duct, and a rear filter arranged in sequence along the air flow direction; wherein the sensors include multiple piezoelectric sensors and a low-frequency vibration sensor, the multiple piezoelectric sensors are arranged at the air inlet, the middle air duct, and the rear filter to collect vibration signals of the corresponding positions; and the low-frequency vibration sensor is arranged at the holding portion to collect low-frequency vibration signals of the holding portion.

[0072] The specific steps of synchronously collecting vibration signals of at least two key positions in the air duct system 06 and vibration signals of the holding portion through multiple sensors include: synchronously collecting vibration signals of the air inlet, the middle air duct, and the rear filter through the multiple piezoelectric sensors, and corresponding numbered as P1 signal, P2 signal, and P3 signal; collecting low-frequency vibration signals of the holding portion through the low-frequency vibration sensor, and numbered as P4 signal.

[0073] It can be understood that the air duct system 06 includes an air inlet, a middle air duct, and a rear filter arranged in sequence along the air flow direction. In this embodiment, the system realizes synchronous monitoring and identification of the internal state of the air duct system 06 and external operation disturbance through multiple sensors arranged at key positions of the dust collector. Specifically, piezoelectric sensors are installed at the air inlet, the middle air duct, and the rear filter to collect vibration signals of different sections in the air flow path; at the same time, a low-frequency vibration sensor is arranged at the holding portion to specifically capture low-frequency mechanical vibrations introduced by user operation. These sensors collect signals with high synchronicity and are numbered as P1, P2, P3, and P4, respectively, to constitute the basic data source of the multi-source perception layer.

[0074] The air inlet piezoelectric sensor (P1) is mainly used for monitoring the collision of the air inlet and the pipeline opening and the medium-high frequency vibration signal excited by the friction between the brush and the ground. Its response characteristic is more sensitive to the initial blockage and the suction of large particles. The air duct piezoelectric sensor (P2) is responsible for capturing the stability change of the airflow in the middle section of the air duct and the wide frequency vibration generated by the impact of solid impurities on the pipe wall, which can effectively identify the blockage or airflow turbulence state in the middle section of the air duct. The piezoelectric sensor (P3) behind the filter screen is located behind the filter screen and directly senses the sound and vibration signals of the airflow after being regulated by the filter screen. Its frequency spectrum characteristic is closely related to the permeability state of the filter screen and can sensitively reflect the blockage degree of the filter screen. The holding part low-frequency vibration sensor (P4) focuses on the low-frequency vibration in the frequency band of 20-200Hz, which is used to extract the contact state of the brush with different types of ground and the interference characteristics introduced by the user's dragging, swinging and other human operations, so as to effectively distinguish the real fault from the operation noise.

[0075] After the system is started, the initialization configuration of each sensor module and its signal conditioning circuit is first completed to ensure that the sampling rate is uniform, the channels are synchronized and the signal quality is stable. Then the real-time data acquisition is started, and the original voltage signals of P1 to P4 are synchronously acquired. The specific implementation process is as follows: the system collects the vibration signal at the air inlet position through the piezoelectric sensor arranged at the air inlet position, which is recorded as P1 signal; the vibration signal at the middle part of the air duct is collected through the piezoelectric sensor at the middle part of the air duct, which is recorded as P2 signal; the vibration response at the rear of the filter screen is collected through the piezoelectric sensor at the rear of the filter screen, which is recorded as P3 signal; at the same time, the low-frequency vibration signal at the holding part is collected through the low-frequency vibration sensor installed at the holding part, which is recorded as P4 signal. The four signals are kept time-aligned through high-precision synchronous sampling mechanism, providing reliable data basis for subsequent signal analysis and state recognition.

[0076] The traditional scheme usually relies on high-cost air pressure or flow sensors, which not only have complex installation, but also are easily disturbed and have low reliability in complex working conditions. In contrast, the present application uses low-cost piezoelectric sensors combined with multi-node synchronous sampling strategy and introduces blind source separation signal processing method, which can decouple different sources of fault characteristics from mixed vibration signals, thereby realizing high cost-effective blockage discrimination. In particular, by adding a holding part low-frequency vibration sensing channel, the system can effectively identify and filter out the low-frequency disturbance caused by user operation, improving the accuracy of blockage judgment and the robustness of the system in real use scenarios.

[0077] In an embodiment, the step S100 of signal processing on the collected vibration signal to extract the characteristic value representing the airflow state inside the air duct system 06 and the characteristic value representing the contact state of the brush of the vacuum cleaner with the ground includes steps S210-S220:

[0078] S210: pre-processing the collected vibration signal to obtain a pre-processed vibration signal;

[0079] S220: performing time-frequency transformation on the pre-processed vibration signal to obtain a corresponding frequency domain signal;

[0080] S230: extracting a feature value for representing the airflow state inside the air duct system 06 and a feature value for representing the contact state between the floor brush and the ground from the frequency domain signal.

[0081] It can be understood that, in one specific embodiment of the present application, step S100 involves in-depth processing of the collected multi-channel vibration signal to extract key feature values that can effectively represent the airflow state inside the air duct system 06 and the contact state between the floor brush and the ground, specifically including three core steps of signal preprocessing, time-frequency transformation and feature extraction, the process of which is shown in FIG. 2. It is worth noting that in the present application, “P1”, “P2”, “P3” and “P4” refer to the physical sensor itself, i.e. “P1” and “P1 sensor” are different descriptions of the same meaning, and “P1 signal” represents the signal collected and output by “P1”. Figure 2

[0082] Step S210 is the signal preprocessing stage, which aims to purify the original signal and suppress noise to provide high-quality data for subsequent analysis. This step implements differentiated filtering strategies according to the characteristics and physical meanings of different sensor signals: for the P4 signal of the holding part low-frequency vibration sensor, low-pass filtering is adopted to retain pure low-frequency vibration components, effectively separating the low-frequency disturbance introduced by the interaction between the floor brush and the ground and the user's operation; for the P1 signal, P2 signal and P3 signal of the piezoelectric sensors at the air inlet, the middle of the air duct and behind the filter screen, band-pass filtering is implemented, for example, 50Hz-5kHz, to suppress the motor 03 power frequency interference and high-frequency noise, while retaining the medium-high frequency vibration characteristics related to airflow dynamics and blockage.

[0083] It is worth noting that the present embodiment introduces a blind source separation (BSS) algorithm in this step. Blind source separation estimates and separates the independent source signals hidden therein only according to the statistical characteristics of multiple observation signals without any prior knowledge of the source signals. In our application scenario, the signals collected by each sensor are the results of multiple vibration sources mixed together. For example: the P1 signal may contain airflow impact noise, particle collision sound and low-frequency structure transmission vibration from the floor brush; the P4 signal may contain working vibration from the floor brush and shaking noise from the user's hand during operation. Traditional filtering methods can only perform rough screening according to the frequency range, and cannot distinguish physical events with different sources in the same frequency band. Blind source separation can fundamentally solve this problem.

[0084] ​The implementation of BSS in S210 is as follows: in the preprocessing sub-step, the system inputs the multiple mixed signals (P1-P4) collected synchronously into the BSS algorithm, and finally outputs several independent components (ICs) through iterative optimization. These components are the separated and estimated source signals.

[0085] The separation results can be as follows: IC1: a component mainly containing airflow dynamic noise, which can be mainly decomposed from P2 and P3 signals; IC2: a component mainly containing transient impact of solid matter collision and friction, which can be mainly from P1 and P2; IC3: a component mainly containing low-frequency vibration conduction of the brush motor 03, which can be clearly present in P4 and IC2; IC4: a component mainly containing user operation interference, which can be uniquely embodied in P4.

[0086] After introducing BSS, the feature purity is greatly improved, and the objects processed in the subsequent steps S220 of FFT transformation and S230 of feature extraction are changed from the original "mixed soup" signal to these independent components (ICs) with clear physical meaning. For example, the features representing airflow can be extracted from the "purified" IC1, and the features representing inhalation collision can be extracted from IC2, so as to greatly reduce the cross interference between different physical phenomena. At the same time, the anti-interference ability is enhanced, and in the subsequent identification, the system can analyze the components (IC1, IC2, IC3) representing the internal state or directly ignore the interference component (IC4), so as to clearly distinguish the real blockage from the false abnormality caused by human operation, and greatly reduce the false alarm rate.

[0087] Each independent component is highly related to a physical source, which makes the mapping relationship between the feature values based on these components, such as the entropy of IC1 and the energy of IC2, and the actual working state of the system more direct and clear, facilitating the debugging and optimization of the algorithm. In summary, the introduction of blind source separation in step S210 is no longer simply shaping the signal, but actively and intelligently decomposing the mixed signal into its physical causes, laying a pure and reliable data foundation for subsequent time-frequency analysis and feature extraction, which is the core of the entire scheme to realize high-precision and high-robustness state recognition.

[0088] Step S220 performs time-frequency transformation on the pre-processed time-domain vibration signals. Each route time-domain signal is converted into a frequency-domain signal through Fast Fourier Transform (FFT), thereby obtaining a frequency spectrum that can clearly reflect the frequency structure and energy distribution of the signal. This conversion can reveal the frequency composition and energy distribution of the vibration signal, and is a key step in identifying state-related characteristic frequency components. As an efficient algorithm for calculating Discrete Fourier Transform (DFT), the core of FFT is to decompose complex time-domain waveforms into a series of sine waves with different frequencies, amplitudes, and phases. For each pre-processed signal, the FFT calculation outputs a complex number array, and the amplitude spectrum directly shows the energy intensity of the signal at each frequency point. The generated frequency spectrum has a horizontal axis representing frequency and a vertical axis representing amplitude or power, thereby clearly showing features such as periodic components, resonance peaks, and harmonics in the time domain that are difficult to directly observe.

[0089] Step S230 extracts characteristic values representing the airflow state inside the air duct system 06 and representing the contact state of the brush of the vacuum cleaner with the ground from the frequency-domain signals. This step converts the frequency-domain signals obtained after preprocessing and time-frequency transformation into numerical indicators that can quantify the key working states of the vacuum cleaner. The composition of this feature is not a single parameter, but a set of "feature vectors". This set usually contains two types of core information: one is the feature for describing the health of the airflow inside the air duct system 06, such as indicators reflecting the degree of airflow turbulence, smoothness, or blockage; the other is the feature for judging the physical contact quality of the brush with the ground, such as indicators reflecting whether the brush is suspended or effectively scraping the ground.

[0090] Based on the core principle that different physical states leave unique "fingerprints" on the vibration spectrum. Under different airflow states such as normal suction, partial blockage, and complete blockage, the frequency distribution, energy concentration, and harmonic components of the acoustic and vibration signals induced by the airflow will show regular changes. Similarly, when the brush effectively contacts the ground, is lifted, or moves on rough / slippery ground, the mechanical vibration spectrum characteristics generated by the brush are also completely different. The working principle of this step is to capture and condense these abstract spectral differences into specific, calculable numerical values through algorithms such as calculating frequency band energy, spectral entropy, amplitude ratio, and main frequency component. These numerical values form a feature space, and different states will gather in different regions in this space, thereby achieving accurate differentiation.

[0091] Through the above processing flow, the system can extract characteristic values with clear physical meaning from multiple vibration signals, providing a reliable information foundation for subsequent accurate identification of the working state of the vacuum cleaner and differentiation between real blockage and operational interference.

[0092] In an embodiment, the specific steps of extracting feature values for characterizing the airflow state inside the air duct system 06 and the feature values for characterizing the contact state of the dust cleaner's floor brush with the ground from the frequency domain signals include:

[0093] From the frequency domain signal of the P1 signal, a first feature value is extracted, which is used to characterize the airflow state at the air inlet; from the frequency domain signals of the P2 signal and the P3 signal, a second feature value is extracted, which is used to characterize the balance state of the air duct pressure; from the frequency domain signal of the P4 signal, a third feature value is extracted, which is used to characterize the contact state of the dust cleaner's floor brush with the ground.

[0094] It can be understood that the embodiment is a specific implementation scheme of the previous embodiment, which assigns specific sensor signal sources and specific calculation targets to the three key physical states, forming a multi-dimensional and cross-verified perception system. By utilizing different information perceived by sensors at different positions on the dust cleaner body and through collaborative analysis, the accuracy and robustness of state recognition are greatly improved.

[0095] From the frequency domain signal of the P1 signal, a first feature value is extracted, which is used to characterize the airflow state at the air inlet. The first feature value is the air inlet signal entropy value (H1) S1 , which is obtained by calculating the information entropy of the P1 signal in the main frequency band, such as 100 Hz-2 kHz. The entropy value represents the complexity and randomness of the signal: a high entropy value corresponds to a complex and highly random signal, which characterizes the normal suction state of the airflow; a low entropy value means a simple and periodic signal, indicating that the air inlet may be blocked.

[0096] From the frequency domain signals of the P2 signal and the P3 signal, a second feature value is extracted, which is used to characterize the balance state of the air duct pressure. The second feature value is the air duct and filter screen signal amplitude ratio (R2,3) R , which is obtained by calculating the amplitude ratio of the P2 and P3 sensors at a specific characteristic frequency point or in a specific frequency band. R = Amp P2 / Amp P3

[0097] From the frequency domain signal of the P4 signal, a third feature value is extracted, which is used to characterize the contact state of the dust cleaner's floor brush with the ground. The third feature value is the handle low-frequency energy (E4) E low , which is obtained by calculating the integral energy of the P4 signal in the 20-200 Hz frequency band. This feature directly reflects the contact state of the floor brush with the ground and the effectiveness of the cleaning action. High energy usually indicates that the floor brush is working normally, and low energy may indicate that the floor brush is suspended or not in good contact.​​​​

[0098] In an embodiment, the second characteristic value is extracted from the frequency domain signals of the P2 signal and the P3 signal, and the specific steps for representing the balance state of the air duct pressure include:

[0099] The first amplitude data in the first predetermined frequency point or the first narrow frequency band is extracted from the frequency domain signal of the P2 signal, and the second amplitude data corresponding to the first predetermined frequency point or the first narrow frequency band is extracted from the frequency domain signal of the P3 signal; the ratio of the first amplitude data and the second amplitude data is calculated to obtain the second characteristic value representing the balance state of the air duct pressure.

[0100] It can be understood that, in the embodiment, the second characteristic value representing the balance state of the air duct pressure is extracted from the frequency domain signals of the P2 signal and the P3 signal, and the specific steps are as follows:

[0101] The first amplitude data in the first predetermined frequency point or the first narrow frequency band is extracted from the frequency domain signal of the P2 signal, which directly reflects the vibration intensity of a specific source inside the air duct. The P2 sensor is located on the wall of the air duct, and its signal directly senses the airflow pressure pulsation and solid vibration inside the air duct. The airflow in the air duct is not smooth and silent, but full of pressure pulsations of specific frequencies generated by the rotation of the fan and the turbulence. These pressure pulsations will excite the vibration of the same frequency on the wall of the air duct. The vibration amplitude is positively correlated with the pressure amplitude that excites it. Therefore, by measuring the vibration amplitude of the P2 sensor signal at the characteristic frequency, the dynamic pressure level at the point in the air duct can be indirectly reflected.

[0102] The purpose of selecting the "first predetermined frequency point" or "first narrow frequency band" is to accurately lock a characteristic frequency component that is most sensitive to the change of the airflow state in the air duct and stable. This characteristic frequency may be determined by the blade passing frequency of the fan, the resonance frequency of the air duct cavity, or other specific phenomena strongly related to airflow dynamics. The amplitude of the specific frequency component, i.e., the first amplitude data, is extracted as an index representing the absolute pressure or vibration intensity at the inlet end of the air duct. When the air duct is blocked, the pressure at this point will rise, resulting in an increase in the vibration amplitude of the characteristic frequency. The "first amplitude data" refers to the energy intensity or amplitude of the vibration signal in a specific frequency point or a very narrow frequency band range on the P signal spectrum obtained by fast Fourier transform, which is determined by pre-experiment or theoretical analysis.

[0103] From the frequency domain signal of the P3 signal, a second amplitude data corresponding to the first predetermined frequency point or the first narrow frequency band is extracted to capture the attenuation of the same vibration source intensity behind the filter screen. The P3 sensor is located behind the filter screen, and it senses the vibration and residual pressure pulsation that penetrates through the filter screen. The same source vibration generated by the fan will be affected by the filter screen material, the dust accumulation in the dust bucket, and the degree of smoothness of the downstream pipeline when propagating downstream to the air duct. The filter screen and dust act as an acoustic and vibration "damper" and "filter", which absorbs and reflects part of the energy. Therefore, the amplitude measured at the P3 sensor reflects the degree of attenuation of the vibration of the characteristic frequency from the source point to the measurement point. The greater the downstream resistance, the more serious the attenuation, and the smaller the amplitude of the P3 sensor.

[0104] The amplitude at the same frequency point or frequency band as P2, i.e. the second amplitude data, is extracted, which aims to establish a benchmark for direct comparison with P2. Its role is to represent the vibration intensity of the same physical source measured behind the filter screen. When the filter screen is blocked or the downstream of the air duct is blocked, the vibration and pressure wave will be more hindered in the propagation process, resulting in a decrease in the amplitude of the frequency component measured at P3. Like the "first amplitude data", the "second amplitude data" is extracted from the frequency spectrum of the P3 signal at the same "first predetermined frequency point or first narrow frequency band". Ensuring that the two data strictly correspond in frequency is a prerequisite for subsequent effective comparison and calculation.

[0105] The ratio of the first amplitude data to the second amplitude data is calculated to obtain the second characteristic value representing the balance state of the air duct pressure. Directly comparing the absolute amplitudes of P2 and P3 is susceptible to interference, because changes in fan speed or power fluctuations will simultaneously cause the amplitudes of P2 and P3 to change in the same direction. However, calculating the ratio of the two, the second characteristic value = P2 amplitude / P3 amplitude, can offset the common change in source intensity, thereby isolating the relative attenuation of the vibration on the path from P2 to P3. This ratio is a dimensionless number that is extremely sensitive to the blocked position of the air duct system 06 and can clearly distinguish between a blockage at the front end of the air duct and a blockage behind the filter screen.

[0106] When the air duct system 06 is smooth, the propagation path of the vibration from P2 to P3 is relatively smooth, and the attenuation is small. Therefore, although the amplitudes of P2 and P3 will change with the fan power, their ratio will stabilize within a certain "baseline" range. When the front end of the air duct is blocked, the pressure at P2 will abnormally increase, while the pressure at P3 may decrease due to reduced flow. This increase and decrease will cause the ratio to increase. When the filter screen or the downstream is blocked, the pressure at P2 will also increase due to the increase in back pressure, but the vibration will be severely hindered in the process of propagating to P3, resulting in a sharp decrease in the second amplitude. This combined effect will cause the ratio to increase sharply to a higher level.

[0107] In an embodiment, the extracted feature values are fused with the real-time operating current of the motor, and input into a preset identification model for analysis, to identify the specific steps of the current operating state of the vacuum cleaner, comprising:

[0108] Obtain the working current change condition of the motor 03; if the third feature value is less than the first preset threshold value, and the first feature value is less than the second preset threshold value, it is determined that the air inlet of the vacuum cleaner is blocked; if the second feature value is greater than the third preset threshold value, and the working current of the motor 03 continues to slowly rise, it is determined that the filter screen of the vacuum cleaner is blocked; if the third feature value is not less than the first preset threshold value, the first feature value is not less than the second preset threshold value, and the working current of the motor 03 sharply rises and exceeds the preset rated value, and the frequency domain signals of the P2 signal and the P3 signal are abnormal, it is determined that the exhaust of the vacuum cleaner is blocked.

[0109] The present embodiment combines the abstract feature values extracted in the previous steps with the key electrical parameter of the motor 03 current, and accurately diagnoses and distinguishes several typical faults that may occur in the vacuum cleaner through a preset rule "preset identification model" based on a large amount of experimental experience.

[0110] Obtain the working current change condition of the motor 03, the role of this step is to introduce an independent vibration sensor, a crucial electrical load feedback signal. The motor 03 current is the most direct manifestation of the working load of the vacuum cleaner. It is used as a key auxiliary criterion to verify the fault possibility implied by the vibration features, and can identify some specific faults that cannot be judged alone by the vibration sensor. The input of this step is the real-time driving current signal of the motor 03, and the output is the analysis result of the current signal, which usually includes its instantaneous value, the change trend relative to the rated value, and whether it exceeds the preset safety threshold, which is determined based on the actual situation of the motor. The load of the fan motor 03 of the vacuum cleaner is directly related to the airflow resistance. When the air duct is unobstructed, the motor 03 operates with normal current. When a blockage occurs, the airflow resistance increases, and the fan needs to do more work to maintain the speed, resulting in an increase in the load of the motor 03 and the working current. Therefore, the change pattern of the current is an important basis for judging the severity and nature of the blockage.

[0111] When the third feature value (P3) E low ) is less than the first preset threshold value T1, and the first feature value (P1) S1 ) is less than the second preset threshold value T2. This logical rule is specifically used to identify faults caused by the air inlet being completely blocked by large foreign objects such as socks and plastic bags. E low <T1 indicates that there is vibration, the brush is in good contact with the ground, and the possibility of the brush being lifted or suspended is excluded; S1T2 indicates that the spectrum entropy is extremely low, and there is no turbulent noise. The airflow sound signal at the air inlet is abnormally weak or disappears. Normally, there should be strong airflow sound when the floor brush is in good contact. If the airflow sound feature at the air inlet disappears at this time, a strong logical contradiction is formed, and the only reasonable explanation is that the airflow is completely blocked at the inlet at the front end. The motor 03 current may only have a slight increase or remain unchanged in this state, because after the airflow is completely blocked, the fan is actually idling or has very small load. Therefore, it is determined that the air inlet of the dust collector is blocked.

[0112] When the second characteristic value (P2 / P3) R is greater than the third preset threshold T3, and the working current of the motor 03 continues to slowly rise. This logical rule is used to identify the performance slowly decaying fault caused by the filter screen being gradually covered with dust, which is a gradual blockage. R T3 indicates that the upstream and downstream pressures of the air duct are unbalanced, and the vibration signal behind the filter screen is abnormally weak relative to the air duct signal, indicating that the blockage occurs at the filter screen or after it. The motor 03 current continues to slowly rise, which is a time dimension trend judgment, which completely matches the physical process of the filter screen being gradually covered with dust and the slow deterioration of the air permeability. The slowly rising current confirms that the system resistance is continuously increasing. The filter screen blockage is a slow process. The abnormal increase of the vibration feature ratio R indicates the approximate location of the blockage. And the dynamic trend of the motor 03 current "continuously slowly rising" is the most critical point of differentiation from "instantaneous blockage of the air inlet" or "instantaneous blockage of the air outlet", clearly depicting the process of gradually increasing resistance over time. Therefore, it is determined that the filter screen of the dust collector is blocked.

[0113] When the third characteristic value E low is not less than T1, the first characteristic value S1 is not less than T2, and the working current of the motor 03 sharply rises and is greater than the preset rated value, and the frequency domain signals of P2 and P3 signals appear severe abnormalities. This logical rule is used to identify the most dangerous fault, which is the blockage of the air outlet, which can cause the motor 03 to be severely overloaded and have the risk of burning out. The complex determination of multiple conditions needs to be met at the same time: the air inlet and the floor brush are in normal state, excluding the possibility of front blockage; the motor 03 current sharply rises, indicating that the resistance increases sharply in an instant, far exceeding the mode of slow blockage of the filter screen; P2 / P3 spectrum is severely abnormal, and the air outlet blockage can cause the pressure of the entire air duct system 06 to rise sharply, causing strong aerodynamic noise and structural resonance, which appears as amplitude abnormally high or new howling frequency in the spectrum of P2 and P3.

[0114] When the front end is normal but the system resistance is huge, the problem must be in the rear. The "sharp" rise in current contrasts sharply with the "slow" rise in filter clogging, indicating a sudden and serious blockage. The "severe abnormality" of the vibration signal reflects the extreme state of the system's internal pressure. These three evidence chains combined together point to the conclusion that the exhaust system is severely blocked, and the system can take immediate protective measures such as forced speed reduction.

[0115] All the above fault conditions are not met, indicating that the system is currently not detecting any known fault mode and is in a healthy working state.

[0116] In an embodiment, according to the identified current working state, the specific steps of adjusting the working state of the motor 03 include:

[0117] When it is determined that the air inlet of the vacuum cleaner is blocked, the control motor 03 is operated in an intermittent super-power mode, and the user is prompted; when it is determined that the filter of the vacuum cleaner is blocked, the power of the motor 03 is reduced, and the user is prompted; when it is determined that the exhaust of the vacuum cleaner is blocked, the power of the motor 03 is reduced to a safe gear, and a sound and light alarm is issued and the user is prompted.

[0118] This embodiment describes a closed-loop control system based on the state recognition result. According to different types of faults, the working state of the motor 03 is intelligently and differentially adjusted to achieve three major goals: attempt to solve the problem autonomously; protect the safety of the equipment; and guide the user to intervene in the appropriate way. According to the physical causes and potential risks of different faults, the optimal and safest motor 03 control instructions are taken, and clear communication with the user is carried out through the human-machine interface (HMI).

[0119] When it is determined to be "normal" state, the motor 03 continues to work at the current gear. The role of this strategy is to ensure that the vacuum cleaner can continuously provide the user with the cleaning performance expected in a healthy state without any unnecessary intervention, ensuring the continuity of the user experience. The output of this strategy is the "maintain current PWM (pulse width modulation) signal" instruction sent to the motor 03 driver. The human-machine interface (HMI) has no abnormal prompt. The system continues to monitor various characteristic values, and since they are all within the normal range, the control loop performs the "hold" operation, the motor 03 controller maintains the existing power output, and the vacuum cleaner runs at the best performance.

[0120] When the system determines that the "intake is blocked", it controls the motor 03 to run at intermittent over-power (trigger the strong suction pulse mode) and prompts the user on the HMI. Automatically clear the blockage, by the instantaneous strong suction to suck or blow away the foreign matter (such as paper, plastic bag) stuck in the air inlet. Secondly, when it cannot be cleared, it clearly prompts the user to manually check the air inlet. Send a series of short, high-intensity power pulse instructions to the motor 03 driver. Each pulse cycle contains a short period of higher-than-normal power driving phase, followed by a short period of returning to normal or low power. Display the "intake is blocked" icon or text prompt on the display screen or APP, possibly accompanied by a prompt sound.

[0121] Intermittent rather than continuous over-power operation is to prevent the motor 03 from overheating due to long-term overload. During each pulse interval, the system will re-detect the characteristic value. If the blockage is cleared and the characteristic value returns to normal, the system will exit the mode; if the blockage remains after multiple pulses, the user will be prompted to wait for manual processing.

[0122] When the system determines that the "filter is blocked", it reduces the power of the motor 03, such as one to two gears, and gently prompts the user to clean the filter on the HMI. The core of this strategy is to balance between protecting the equipment and maintaining basic functions, and actively reminding the user to maintain. Reducing power can slow down the speed of the filter being further compacted and prevent the motor 03 from being inefficient and overheating due to continuous medium to high load. The gentle prompt aims to inform the user that routine maintenance is needed, rather than an emergency failure. Send instructions to the motor 03 driver to permanently reduce its power output by a preset gear. The "please clean the filter" icon is continuously displayed on the HMI, possibly accompanied by a soft intermittent prompt sound, which will not cause the user to be nervous.

[0123] Filter blockage is a gradual process and is not instantaneous. At this time, the system resistance increases, and the motor 03 runs inefficiently and consumes electricity at the original high power. Active power reduction is a "derated operation" strategy that reduces the overall air volume of the fan, thereby reducing the resistance through the dirty and blocked filter, allowing the motor 03 to work in a lighter and safer load range. This not only protects the motor 03, but also provides buffer time for the user to clean the filter.

[0124] When the system determines that the "exhaust is blocked", it immediately reduces the power of the motor 03 to the lowest safe gear and issues an audible and visual alarm to prompt the user to check immediately. Prevent the motor 03 from burning out in a short time due to extreme overload. Send the highest priority instruction to the motor 03 driver to instantly drop the power to the absolute minimum gear. The red warning icon on the display screen is always on or flashing, emitting a sharp and high-frequency beep, and displaying clear text information.

[0125] Exhaust blockage leads to complete closure of the air duct, the motor 03 load increases sharply, the current surges, the coil temperature will grow exponentially, and it can cause permanent magnet demagnetization or coil burnout in a very short time. Immediately reduce to the minimum safe power, forcibly pull the operating point of the motor 03 away from the danger zone, and save it from the edge of burning out. The sharp alarm is to ensure that the user can immediately notice this serious failure and take physical measures such as shutdown, troubleshooting, etc.

[0126] In an embodiment, the dust collector comprises a pneumatic adjusting mechanism 05 arranged in the air duct system 06; the dust collector control method further comprises:

[0127] S500: According to the identified working state, control the pneumatic adjusting mechanism 05 to act to adjust the airflow state of the air duct system 06. According to the accurate result of S400 state identification, a specific action instruction is issued to the pneumatic adjusting mechanism 05. Intervene in the airflow dynamics environment inside the air duct to achieve the functions of clearing blockage, optimizing the air duct, or performing specific cleaning tasks, etc.

[0128] As shown in Figure 3 , the dust collector control method of the embodiment comprises:

[0129] S100: Synchronously collect vibration signals of at least two different parts in the air duct system 06 and vibration signals of the holding part through multiple sensors;

[0130] S200: Process the collected vibration signals to extract feature values representing the airflow state inside the air duct system 06 and the contact state of the brush of the dust collector with the ground;

[0131] S300: Based on the extracted feature values and the working current of the motor 03, analyze through a preset identification model to identify the current working state of the dust collector;

[0132] S400: Adjust the working state of the motor 03 according to the identified working state.

[0133] S500: According to the identified working state, control the pneumatic adjusting mechanism 05 to act to adjust the airflow state of the air duct system 06.

[0134] Steps S100-S400 are exactly the same as the technical effects of all technical solutions of all embodiments of the above-mentioned dust collector control method, and will not be repeated here.

[0135] The pneumatic adjustment mechanism 05 generally includes an actuator such as a small motor 03, a stepper motor 03 or an electromagnetic solenoid. It is responsible for providing the power required for mechanical movement; an actuator such as a valve, a baffle, a flap or a movable nozzle. It is a mechanical component that directly contacts the air flow and changes its path or state; control circuit, receiving low-power signals from the system, and driving the actuator for precise movement. In step S500, the system calls the corresponding control strategy from the pre-programmed program according to the identified state such as "air inlet blockage". Then, the system generates the corresponding control signal and transmits it to the actuator through the driving circuit. The actuator generates linear or rotary motion according to the instruction, and then drives the actuator to the predetermined position, thereby changing the physical structure of the air duct. This structural change will immediately affect the air flow path, pressure distribution and flow rate in the air duct, achieving active processing of the fault.

[0136] In combination with the above embodiments and the accompanying Figure 4 , the present embodiment proposes the following specific steps including:

[0137] System initialization and synchronous data acquisition, hardware preparation for the entire intelligent control system, and obtaining raw data reflecting the true state of the system. After the vacuum cleaner is powered on, the system first initializes and configures each hardware module to ensure that the sensors and driving circuits are in a ready state. Subsequently, the system continuously collects the raw voltage signals (P1, P2, P3, P4) of the four sensors and the working current signal of the motor 03 in a synchronous sampling manner. Synchronous sampling ensures the alignment of different sensor data in time, laying a foundation for subsequent accurate correlation analysis.

[0138] Signal preprocessing and frequency domain conversion: low-pass filter the P4 signal to extract pure low-frequency vibration components, which best reflect the mechanical contact state of the brush and the ground. Band-pass filter the P1, P2, P3 signals to effectively filter out low-frequency motor 03 structure vibrations and high-frequency electronic noise, retain the key frequency bands that can reflect the air flow state, and introduce a blind source classification algorithm.

[0139] Perform Fast Fourier Transform (FFT) on the preprocessed time domain signals to obtain clear feature spectrum graphs. Frequency domain analysis can separate different frequency components mixed together, facilitating subsequent feature extraction and identification. Extract quantifiable and distinguishable key feature indicators from the purified signals to convert complex sensor data into simple decision-making basis.

[0140] Features E low (Handle low-frequency energy), calculate the integral energy of the P4 signal in the 20-200Hz frequency band. This value directly quantifies the working efficiency of the brush, and a low value indicates that the brush is suspended or not in contact with the ground.

[0141] FeaturesS1 ((Intake port signal entropy value)), calculate the information entropy of the P1 signal within the main frequency band (such as 100 Hz - 2 kHz). The entropy value measures the randomness and complexity of the signal. A high entropy value represents a large variety of foreign objects inhaled by the air flow and a rich sound signal; a low entropy value represents stagnant air flow and a monotonous signal.

[0142] Feature R ((Air duct pressure difference ratio)), calculate the amplitude ratio of the P2 and P3 sensors at specific characteristic frequency points (such as the fundamental frequency harmonic of motor 03). R = Amp ( P2 ) / Amp ( P3 ). This ratio sensitively reflects the pressure difference change on both sides of the filter screen. When the filter screen is blocked, the P3 signal weakens and the ratio R increases significantly.

[0143] The block type recognition system based on the decision tree compares the extracted feature values with the preset thresholds calibrated through a large number of experiments, and comprehensively considers the change trend of the current of motor 03 to execute a rigorous decision tree logic. Intake port blockage criterion: ( E low <T1) and ( S1 <T2). That is, when the floor brush is working normally (not suspended), but there is no air flow sound at the intake port, it can be determined that the blockage occurs at the intake port. Filter screen blockage criterion: ( R >T3) and (the current of motor 03 continuously rises slowly). The air duct pressure difference ratio is abnormal, and at the same time, the load of motor 03 steadily increases, which is a typical feature of the filter screen being gradually blocked by dust. Exhaust blockage criterion: ( E low and S1 normal) but (the current of motor 03 soars sharply). The intake port works normally, but the load of motor 03 is extremely large. The only explanation is that a serious blockage occurs at the exhaust end (dust collection bin, rear filter screen). Normal state: None of the above conditions are met.

[0144] For different faults, execute the most optimized processing strategy, taking into account both the blockage clearing efficiency and equipment safety. In the normal state, maintain the current power, without prompting, to provide the best user experience. When the intake port is blocked, start the strong suction pulse mode (intermittent over - power), try to clear foreign objects with strong suction; prompt the user; optionally start the reverse blowing function for active cleaning. When the filter screen is blocked, reduce the power of motor 03 by 1 - 2 gears to reduce the wind pressure, protect motor 03 and delay blockage; gently prompt the user to clean the filter screen and give a buffer time. When the exhaust is blocked, immediately reduce the power断崖式 to the safe gear, which is the highest - priority protection instruction. The primary purpose is to prevent motor 03 from being instantly burned out; at the same time, trigger the highest - level sound and light alarm, requiring the user to intervene immediately.

[0145] After the control strategy is executed, the flow immediately jumps back to the signal acquisition step to start a new round of data acquisition, analysis and judgment. If the blockage has been cleared and the characteristic value returns to normal, the system automatically exits the special mode and resumes normal operation. If the problem persists, the control strategy is continuously adjusted or adjusted. This closed loop ensures the real-time, adaptability and reliability of the control.

[0146] In addition, the application also proposes a control system, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the above-mentioned dust collector control method. The memory, the processor and the computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the above-mentioned dust collector control method. A main controller can be used to implement, such as a DSP (Digital Signal Process, digital signal processing chip), FPGA (Field Programmable Gate Array, programmable logic gate array chip), MCU (Microcontroller Unit, microcontroller unit), SOC (System On Chip, system chip) and the like.

[0147] It is worth noting that since the control system of the application is applied to the above-mentioned dust collector control method, the embodiments of the control system of the application include all the technical solutions of all the embodiments of the above-mentioned dust collector control method, and the technical effects achieved are also completely the same, which will not be repeated here.

[0148] In addition, as Figure 5 shown, the application also proposes a dust collector, comprising the above-mentioned control system, and an air duct system 06; a holding part; a motor 03; a plurality of sensors respectively arranged at different positions in the air duct system 06 and on the holding part, for collecting vibration signals of different positions in the air duct system 06 and vibration signals of the holding part; a motor driving module 02, the input end is electrically connected with the output end of the control system, the output end is electrically connected with the motor 03, for receiving the control signal of the control system to adjust the power of the motor 03; a man-machine interaction module 04, electrically connected with the control system, for displaying the working state of the dust collector.

[0149] This embodiment describes a complete dust collector device integrating the aforementioned intelligent control system. The four functional modules of perception, decision-making, execution and interaction are organically integrated into a whole that works cooperatively, thereby realizing closed-loop automation from signal acquisition, intelligent analysis to active control. The composition of the whole machine is a typical mechatronic system, and its working principle is to identify the state based on multi-sensor data fusion, and accordingly drive the execution mechanism and the interaction interface to realize intelligent operation.

[0150] The control system processes all input information, executes complex diagnostic algorithms, and issues all control instructions. It is responsible for implementing all the "methods" and "steps" described in the previous embodiments. It is usually centered on a master micro control unit (master MCU 01), supplemented by necessary memory (such as Flash, RAM), analog-to-digital converters (ADC), and peripheral circuits. On the software level, it is embedded with a complete firmware program containing signal processing, feature extraction, state recognition decision trees, and control logic. The master MCU 01 continuously reads data from the sensors, performs fast Fourier transform (FFT), feature value calculation, and threshold comparison internally. By running the pre-set recognition model (decision tree), it can determine the working state of the vacuum cleaner in real time. Finally, it generates corresponding control signals according to the judgment results and sends them to the motor drive module 02 and the human-computer interaction module 04, and even the pneumatic adjustment mechanism 05, to command them to work cooperatively.

[0151] The air duct system 06 generates suction, contains airflow, separates and collects dust. At the intelligent control level, it is the object of monitoring and regulation, and its internal airflow state directly determines the signal characteristics collected by the sensors. It includes the air inlet, hose, hard pipe, dust collection bin, fan volute, and possibly the pneumatic adjustment mechanism 05. The motor 03 drives the fan impeller to rotate at high speed, generating negative pressure in the air duct. External air carrying dust flows into the air inlet. When the airflow passes through the filter screen, the dust is trapped in the dust collection bin, and the clean air passes through the filter screen, flows through the motor 03 for cooling, and finally exits from the exhaust port. Any blockage at any position will change its internal pressure, flow rate, and acoustic vibration characteristics.

[0152] The handle provides ergonomic hand holding points for easy user operation; the second function is as a mounting platform for the sensor (P4) to collect unique vibration signals from user operation and ground feedback. It is usually made of plastic or metal, with P4 vibration sensors pre-embedded or installed inside, and integrated with switches, gear adjustment buttons, and other human-computer interaction elements. When the user pushes or pulls the vacuum cleaner, the specific vibrations generated by the brush rubbing and colliding with different types of ground are transmitted to the handle through the shaft. The P4 sensor captures these vibrations, and their signals are used to determine the working state of the brush, which is a key basis for distinguishing between air inlet blockage and ground type.

[0153] The motor 03 drives the fan impeller to rotate, providing the source of suction for the entire air duct system 06. In intelligent control, it is also the most important controlled object and load feedback source. It receives drive signals from the motor drive module 02, converts electrical energy into mechanical energy, and drives the impeller to rotate. Its working current changes in real time with the change of load, and this current signal is fed back to the control system, becoming the core electrical parameter for judging the system load state.

[0154] A plurality of sensors (P1, P2, P3, P4) form a sensor array 07, usually including sound pressure sensors and low-frequency vibration sensors, which are precisely installed at the air inlet (P1), air duct wall (P2), filter screen (P3) and holding part (P4) of the air duct system 06 respectively. These sensors convert the mechanical vibration or sound pressure fluctuations captured into analog electrical signals. After amplification and filtering, these signals are converted into digital signals by the ADC of the control system for subsequent frequency spectrum analysis and feature extraction by the MCU.

[0155] The motor drive module 02 receives the weak current control signal from the main control MCU 01, and amplifies and converts it into a strong current drive signal that can directly drive the high-power motor 03 to operate. It usually includes a drive chip and a current sampling circuit, which is used to monitor the motor 03 current in real time and feedback to the main control MCU 01. The main control MCU 01 outputs a low-power PWM control signal to the drive chip, which controls the switching time and duty cycle of the power MOSFET or IGBT in the inverter bridge, thereby accurately adjusting the voltage and current applied to the three-phase winding of the motor 03, and finally achieving precise control of the motor 03 speed and torque.

[0156] The human-machine interaction module 04 clearly communicates the internal working state of the vacuum cleaner, the identified fault information and the operation prompts to the user in the form of display and sound, achieving two-way communication. It includes a display unit, a sound unit and an input unit. According to the instructions issued by the control system, the HMI module will drive different LED light colors to flash, display specific text or icon warnings on the screen, or emit prompt sounds / alarm sounds of different frequencies and rhythms, thereby guiding the user to perform the next operation.

[0157] The above only describes some embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent structural transformation made according to the technical concept of the present application, or direct / indirect application in other related technical fields is included in the patent protection scope of the present application.

Claims

1. A method for controlling a dust collector, applied to a dust collector, the dust collector comprising an air duct system, a handle, a motor and a plurality of sensors, the plurality of sensors being distributed in at least two key positions of the air duct system and the handle, characterized in that, The dust collector control method comprises: Synchronously collecting vibration signals of at least two key positions in the air duct system and a vibration signal of the holding part through the plurality of sensors; Signal processing is performed on the collected vibration signals to extract characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the dust collector with the ground; The extracted characteristic values are fused with the real-time working current of the motor and input into a preset recognition model for analysis to identify the current working state of the dust collector; According to the identified current working state, the working parameters of the motor are adjusted.

2. The dust cleaner control method according to claim 1, wherein The air duct system comprises an air inlet, an air duct middle part and a filter rear part arranged along the airflow propagation direction; wherein the sensors comprise a plurality of piezoelectric sensors and a low-frequency vibration sensor, the plurality of piezoelectric sensors are arranged at the air inlet, the air duct middle part and the filter rear part to collect vibration signals of the corresponding positions; the low-frequency vibration sensor is arranged at the holding part to collect the low-frequency vibration signal of the holding part; the specific steps of synchronously collecting vibration signals of at least two key positions in the air duct system and the vibration signal of the holding part through the plurality of sensors comprise: Synchronously collecting vibration signals of the air inlet, the air duct middle part and the filter rear part through the plurality of piezoelectric sensors, and corresponding numbered as P1 signal, P2 signal and P3 signal; Collecting the low-frequency vibration signal of the holding part through the low-frequency vibration sensor, and numbered as P4 signal.

3. The dust cleaner control method according to claim 2, wherein The specific steps of signal processing on the collected vibration signals to extract characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the dust collector with the ground comprise: Pretreating the collected vibration signals to obtain pretreated vibration signals; Performing time-frequency transformation on the pretreated vibration signals to convert to corresponding frequency domain signals; Extracting characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the dust collector with the ground from the frequency domain signals.

4. The dust cleaner control method according to claim 3, wherein The specific steps of extracting characteristic values representing the airflow state inside the air duct system and characteristic values representing the contact state of the brush of the dust collector with the ground from the frequency domain signals comprise: Extracting a first characteristic value from the frequency domain signal of the P1 signal, the first characteristic value being used to represent the airflow state of the air inlet; Extracting a second characteristic value from the frequency domain signals of the P2 signal and the P3 signal, the second characteristic value being used to represent the balance state of the air duct pressure; Extracting a third characteristic value from the frequency domain signal of the P4 signal, the third characteristic value being used to represent the contact state of the brush of the dust collector with the ground.

5. The dust cleaner control method according to claim 4, wherein The specific steps of extracting a second characteristic value from the frequency domain signals of the P2 signal and the P3 signal, the second characteristic value being used to represent the balance state of the air duct pressure comprise: Extracting first amplitude data in a first predetermined frequency point or a first narrow frequency band from the frequency domain signal of the P2 signal; extracting second amplitude data corresponding to the first predetermined frequency point or the first narrow frequency band from a frequency domain signal of the P3 signal; calculating a ratio of the first amplitude data and the second amplitude data to obtain a second feature value representing a balance state of the wind channel pressure.

6. The dust cleaner control method according to claim 4, wherein The specific step of fusing the extracted feature value with the real-time working current of the motor and inputting into a preset recognition model for analysis to identify the current working state of the dust collector includes: obtaining the working current change condition of the motor; if the third feature value is less than a first preset threshold value, and the first feature value is less than a second preset threshold value, it is determined that the air inlet of the dust collector is blocked; if the second feature value is greater than a third preset threshold value, and the working current of the motor continues to slowly rise, it is determined that the filter screen of the dust collector is blocked; if the third feature value is not less than the first preset threshold value, the first feature value is not less than the second preset threshold value, and the working current of the motor sharply rises and exceeds a preset rated value, and the frequency domain signals of the P2 signal and the P3 signal are abnormal, it is determined that the exhaust of the dust collector is blocked.

7. The dust cleaner control method according to claim 6, wherein The specific step of adjusting the working parameters of the motor according to the identified current working state includes: when it is determined that the air inlet of the dust collector is blocked, the motor is controlled to perform intermittent super-power operation mode, and the user is prompted; when it is determined that the filter screen of the dust collector is blocked, the power of the motor is reduced, and the user is prompted; when it is determined that the exhaust of the dust collector is blocked, the power of the motor is reduced to a safety gear, and an audible and visual alarm is issued and the user is prompted.

8. The method of claim 1-7, wherein, The dust collector comprises a pneumatic adjusting mechanism arranged in the wind channel system; the dust collector control method further comprises: controlling the pneumatic adjusting mechanism to act according to the identified working state to adjust the airflow state of the wind channel system.

9. A control system characterized by, comprises: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the dust collector control method according to any one of claims 1 to 8.

10. A vacuum cleaner comprising: comprises the control system according to claim 9, and a wind channel system; a holding portion; a motor; a plurality of sensors arranged at different positions in the wind channel system and on the holding portion respectively for collecting vibration signals at different positions in the wind channel system and vibration signals of the holding portion; a motor driving module having an input end electrically connected to an output end of the control system and an output end electrically connected to the motor for receiving control signals of the control system to adjust the power of the motor; a man-machine interaction module electrically connected to the control system for displaying the working state of the dust collector.

Images