An abnormality identification system for campus self-service laundry equipment based on the Internet of Things

Abstract

The present invention relates to the technical field of the Internet of Things, and discloses an anomaly identification system for campus self-service laundry equipment based on the Internet of Things. The system comprises: a user operation behavior alignment module, which quantifies the behavior deviation through a dynamic time warping algorithm to generate a first weight factor; a multi-source state feature extraction module, which extracts the device state matrix in combination with vibration spectrum and motor current waveform analysis to generate a second weight factor; a health fusion decision module dynamically allocates weights based on an entropy method, calculates health scores to trigger device-level anomaly detection; an anomaly propagation topology modeling module constructs a fusion topology map, dynamically allocates propagation weights based on historical failure rates and real-time loads, and generates an anomaly propagation path probability map; a spatiotemporal warning visualization module integrates device location and anomaly probability, divides the warning levels into multiple levels, and dynamically displays the risk distribution through a thermal map. Therefore, the present invention significantly improves the detection accuracy and efficiency of anomaly identification of campus self-service laundry equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to an abnormality identification system for campus self-service laundry equipment based on the Internet of Things. Background Art

[0002] Campus self-service laundry equipment commonly encounters the following operational issues: Traditional systems primarily rely on monitoring single device status (e.g., temperature, current), ignoring the impact of user behavior on device health (e.g., frequent mode switching, use during unusual hours), resulting in incomplete anomaly detection. Furthermore, the propagation paths of device anomalies are poorly modeled, making it impossible to predict and prevent cascading failures. For example, when a device shuts down due to a mechanical failure, devices physically connected to it (e.g., shared water supply lines) or communicating with it (e.g., shared network nodes) may fail in a chain reaction. However, existing technologies are unable to analyze these propagation paths, resulting in delayed maintenance responses. Furthermore, existing technologies fail to integrate multi-source data for comprehensive analysis, resulting in insufficient anomaly identification accuracy and a lack of intuitive visual early warning mechanisms. This leads to inefficient operations and maintenance, prolonged user wait times, and a negative user experience. Therefore, there is an urgent need for an intelligent anomaly identification system to improve the efficiency and accuracy of anomaly identification for campus self-service laundry equipment. Summary of the Invention

[0003] The present invention provides a campus self-service laundry equipment anomaly identification system based on the Internet of Things, the main purpose of which is to solve the problem of low efficiency in anomaly identification of campus self-service laundry equipment.

[0004] To achieve the above objectives, the present invention provides a campus self-service laundry equipment abnormality identification system based on the Internet of Things, the system comprising:

[0005] A user operation behavior alignment module is used to generate a user operation time series log of the laundry appliance, align the user operation time series log with a preset standard operation template using a dynamic time warping algorithm, calculate the operation behavior deviation and generate a first weight factor representing the probability of abnormal user behavior;

[0006] a multi-source state feature extraction module, configured to construct a device state matrix of the laundry appliance based on multi-source operation data acquired in real time, and extract characteristic parameters exceeding a preset safety threshold in the device state matrix to generate a second weighting factor representing an abnormal physical state of the appliance;

[0007] a health fusion decision module, configured to perform a weighted decision on the first weight factor and the second weight factor to obtain a health score of the laundry appliance, and trigger device-level anomaly detection when the health score is lower than a first threshold to obtain a device-level anomaly type of the laundry appliance;

[0008] An anomaly propagation topology modeling module is used to construct a fusion topology diagram of the physical connection topology and data communication topology of the laundry device, dynamically assign an anomaly propagation weight to each node in the fusion topology diagram based on the historical failure rate of the device and the real-time load status, and generate an anomaly propagation path probability diagram based on the connection strength between the nodes in the fusion topology diagram and the anomaly propagation weight;

[0009] The spatiotemporal warning visualization module is used to calculate the group anomaly probability of the laundry equipment based on the equipment-level anomaly type and the anomaly propagation path probability map, divide the alarm level into multiple levels according to the probability interval, and integrate the location information of the laundry equipment with the multiple levels of alarm levels to generate a visual alarm log.

[0010] Optionally, the user operation behavior alignment module generates a user operation time sequence log of the laundry device, including:

[0011] Collecting user operation behavior data, wherein the user operation behavior data includes: reservation time period, washing mode selection and payment record;

[0012] A user operation time series log including a reservation period, a washing mode selection sequence and a payment success rate is generated based on the user operation behavior data.

[0013] Optionally, when the user operation behavior alignment module aligns the user operation time series log with a preset standard operation template using a dynamic time warping algorithm and calculates the operation behavior deviation as a first weighting factor of the device health, the module includes:

[0014] Extracting the reservation time period distribution density, mode switching frequency, and payment interruption times from the user operation time series log as key behavior indicators;

[0015] Calculating the cumulative path distance between each of the key behavior indicators and the preset standard operation template using a dynamic time warping algorithm;

[0016] A weighted sum is performed on the accumulated path distances to obtain an operation behavior deviation, and the operation behavior deviation is used as a first weight factor of the device health.

[0017] Optionally, the method for generating a preset standard operation template in the user operation behavior alignment module includes:

[0018] Collecting user operation data of N laundry appliances during a normal operation and maintenance cycle to form a training set;

[0019] Performing Gaussian kernel density estimation on the appointment time period data in the training set to generate a time period distribution probability curve;

[0020] Extract the optimal path of washing mode switching as the standard mode sequence;

[0021] The normal distribution interval of the statistical payment success rate is used as the benchmark payment behavior indicator;

[0022] A preset standard operation template is generated based on the time period distribution probability curve, standard pattern sequence and benchmark payment behavior indicator.

[0023] Optionally, when the multi-source state feature extraction module constructs the device state matrix of the laundry device based on the multi-source operation data acquired in real time, the multi-source state feature extraction module includes:

[0024] Acquire multi-source operation data of the laundry device in real time, wherein the multi-source operation data includes: vibration spectrum data and motor current waveform;

[0025] Performing wavelet packet decomposition on the vibration spectrum data to extract the energy proportion of a preset frequency band as a mechanical wear feature;

[0026] After performing fast Fourier transform on the motor current waveform, the amplitude ratio of the fundamental wave to the third harmonic is calculated as the electrical abnormality feature;

[0027] The mechanical wear features and the electrical abnormality features are aligned according to timestamps to form a two-dimensional state matrix, and the two-dimensional state matrix is ​​determined to be the device state matrix of the laundry device.

[0028] Optionally, when the health fusion decision module performs a weighted decision on the first weight factor and the second weight factor to obtain the health score of the laundry appliance, the module includes:

[0029] Dynamically calculating a weight distribution ratio of the first weight factor and the second weight factor based on an entropy method;

[0030] The health score of the laundry appliance is calculated using a weighted summation algorithm and the weight distribution ratio, wherein the weighted summation algorithm is:

[0031]

[0032] in, is the health score of the laundry appliance, is the weight distribution ratio corresponding to the first weight factor, is the weight distribution ratio corresponding to the second weight factor, is the first weighting factor, is the second weighting factor.

[0033] Optionally, when the abnormality propagation topology modeling module dynamically allocates abnormality propagation weights of each node in the fusion topology graph based on historical equipment failure rates and real-time load status, the module includes:

[0034] Determine the initial static weight of each node according to the betweenness centrality and historical failure rate of the laundry device in the fusion topology graph;

[0035] When the real-time load state of the laundry appliance exceeds 70% of the rated value, determining a dynamic correction coefficient of the initial static weight, and correcting the node weights in the fusion topology graph based on the dynamic correction coefficient;

[0036] When the laundry appliance is in a maintenance lock state, the node weights of the fusion topology graph are forcibly set to zero to block abnormal propagation.

[0037] Optionally, when the spatiotemporal warning visualization module calculates the group abnormality probability of the laundry devices based on the device-level abnormality type and the abnormal propagation path probability graph, the module includes:

[0038] The device-level anomaly type and the anomaly propagation path probability map are input into a spatiotemporal propagation model to calculate the group anomaly probability of the laundry devices, wherein the spatiotemporal propagation model satisfies:

[0039] For mechanical anomalies, an exponential decay propagation rule based on physical distance is adopted;

[0040] For data anomalies, a linear propagation rule based on the number of network hops is adopted;

[0041] When two types of anomalies coexist, mechanical anomaly propagation calculation is performed first and the data anomaly correction term is superimposed.

[0042] Optionally, the calculation formula for the group anomaly probability in the spatiotemporal warning visualization module is as follows:

[0043]

[0044] in, is the population abnormality probability of the laundry equipment, It is due to the initial abnormal device the number of devices that may fail due to the spread of is the number of all laundry appliances in a specific area, is the equipment number of the affected laundry equipment, It is a laundry appliance The exception propagates to the laundry equipment The final probability of is the equipment number of the laundry equipment in a specific area, It's laundry equipment the probability of an anomaly occurring independently (independent of propagation), It's laundry equipment. is the device number of the device that initially failed.

[0045] Optionally, when the spatiotemporal warning visualization module generates a visual alarm log by fusing the location information of the laundry appliance with the multi-level alarm levels, the module includes:

[0046] Rendering a three-level thermal map of abnormal laundry equipment on a campus electronic map based on the location information of the laundry equipment and the multi-level alarm level, where red, yellow, and blue correspond to high risk, medium risk, and low risk, respectively;

[0047] Mark the possible impact range and estimated arrival time of the abnormal propagation of the abnormal device;

[0048] Automatically generate maintenance work orders containing spare parts replacement lists and operation instruction videos based on the pre-associated maintenance knowledge base.

[0049] The present invention uses a dynamic time warping algorithm to align user operation timing with a standard template, quantify behavioral deviations, accurately identify misoperations or malicious behaviors, and avoid equipment failures caused by abnormal user behavior. At the same time, it combines vibration spectrum and motor current waveform analysis to capture mechanical wear and electrical anomalies in real time, realize multi-dimensional physical state monitoring, and enhance fault prediction capabilities. It dynamically allocates weights of user behavior and equipment status through the entropy method to generate a comprehensive health score, which helps to quickly locate high-risk equipment and prioritize critical faults, thereby reducing downtime; based on physical connections and data communication topologies, it predicts abnormal propagation paths, dynamically adjusts node weights, blocks the impact range of high-risk equipment, avoids chain failures, and reduces operation and maintenance costs; in addition, it dynamically displays the abnormal distribution and risk level of campus equipment through thermal maps, and combines location information to mark the abnormal propagation range and expected impact time, assisting managers in quickly formulating emergency strategies, thereby comprehensively improving the efficiency of abnormal identification of campus self-service laundry equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a system architecture diagram of an abnormality identification system for campus self-service laundry equipment based on the Internet of Things provided by one embodiment of the present invention.

[0051] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments belong to some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0053] The terms used in the embodiments of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The singular forms "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates otherwise, and "a plurality" generally includes at least two.

[0054] As used herein, the words “if” or “when” may be interpreted as “at the time of” or “when” or “in response to determining” or “in response to detecting,” depending on the context. Similarly, the phrases “if it is determined” or “if (stated condition or event) is detected” may be interpreted as “when it is determined” or “in response to the determination” or “when detecting (stated condition or event)” or “in response to detecting (stated condition or event),” depending on the context.

[0055] In addition, the step sequence in the following method embodiments is only an example and not a strict limitation.

[0056] In practice, the server-side device deployed in an IoT-based campus self-service laundry equipment anomaly identification system may consist of one or more devices. The aforementioned IoT-based campus self-service laundry equipment anomaly identification system can be implemented as a service instance, a virtual machine, or hardware devices. For example, an IoT-based campus self-service laundry equipment anomaly identification system can be implemented as a service instance deployed on one or more devices in a cloud node. Simply put, an IoT-based campus self-service laundry equipment anomaly identification system can be understood as software deployed on a cloud node, providing a IoT-based campus self-service laundry equipment anomaly identification system to each user terminal. Alternatively, an IoT-based campus self-service laundry equipment anomaly identification system can be implemented as a virtual machine deployed on one or more devices in a cloud node. This virtual machine contains application software for managing each user terminal. Alternatively, an IoT-based campus self-service laundry equipment anomaly identification system can be implemented as a server-side device composed of multiple hardware devices of the same or different types, with one or more hardware devices configured to provide a IoT-based campus self-service laundry equipment anomaly identification system to each user terminal.

[0057] In terms of implementation, the IoT-based campus self-service laundry equipment anomaly identification system and the user end are mutually compatible. Specifically, the IoT-based campus self-service laundry equipment anomaly identification system is an application installed on a cloud service platform, and the user end is a client that establishes a communication connection with the application. Alternatively, the IoT-based campus self-service laundry equipment anomaly identification system is implemented as a website, and the user end is implemented as a webpage. Alternatively, the IoT-based campus self-service laundry equipment anomaly identification system is implemented as a cloud service platform, and the user end is implemented as a mini-program within an instant messaging application.

[0058] like Figure 1 FIG. 1 is a system architecture diagram of an abnormality identification system for campus self-service laundry equipment based on the Internet of Things provided by one embodiment of the present invention.

[0059] The IoT-based campus self-service laundry equipment anomaly identification system 100 described in the present invention can be installed in a cloud server. In terms of implementation, it can be implemented as one or more service devices, as an application installed in the cloud (e.g., a mobile service operator's server, server cluster, etc.), or developed as a website. Depending on the functionality implemented, the IoT-based campus self-service laundry equipment anomaly identification system 100 can include a user operation behavior alignment module 101, a multi-source state feature extraction module 102, a health fusion decision module 103, an anomaly propagation topology modeling module 104, and a spatiotemporal warning visualization module 105. The modules described in the present invention, also referred to as units, refer to a series of computer program segments that can be executed by an electronic device processor and perform fixed functions, and are stored in the electronic device's memory.

[0060] In an embodiment of the present invention, in a campus self-service laundry equipment anomaly identification system based on the Internet of Things, each of the above modules can be implemented independently and called with other modules. The call here can be understood as a module that can connect to multiple modules of another type and provide corresponding services to the multiple modules connected to it. In an campus self-service laundry equipment anomaly identification system based on the Internet of Things provided by an embodiment of the present invention, the scope of application of the campus self-service laundry equipment anomaly identification system architecture based on the Internet of Things can be adjusted by adding modules and directly calling them without modifying the program code, thereby realizing cluster-type horizontal expansion, so as to achieve the purpose of quickly and flexibly expanding a campus self-service laundry equipment anomaly identification system based on the Internet of Things. In actual applications, the above modules can be set in the same device or different devices, or they can be set in virtual devices, such as service instances in cloud servers.

[0061] The following describes the various components and specific workflows of an abnormality identification system for campus self-service laundry equipment based on the Internet of Things in conjunction with specific embodiments.

[0062] The user operation behavior alignment module 101 is used to generate a user operation time series log of the laundry device, align the user operation time series log with a preset standard operation template through a dynamic time warping algorithm, calculate the operation behavior deviation and generate a first weight factor representing the probability of abnormal user behavior.

[0063] In an embodiment of the present invention, generating a user operation time sequence log of a laundry device includes:

[0064] Collecting user operation behavior data, wherein the user operation behavior data includes: reservation time period, washing mode selection and payment record;

[0065] A user operation time series log including a reservation period, a washing mode selection sequence and a payment success rate is generated based on the user operation behavior data.

[0066] In detail, the user operation time series log is a chronological record of the user's behavior data when using the laundry equipment, including reservation time, selected washing mode, payment status and other information, in order to capture user behavior patterns and provide a data basis for subsequent anomaly detection.

[0067] In detail, traditional systems only record device status (such as temperature and current), ignoring the impact of user operations on device health (such as frequent reservations + high water level mode may accelerate device wear).

[0068] In detail, the reservation time period refers to the specific time when the user reserves the use of the washing machine (such as 2023-10-0518:30); the washing mode selection refers to the mode selected by the user (such as "quick wash for 15 minutes" and "high temperature sterilization"); the payment record refers to whether the payment is successful and the payment method (such as the number of WeChat payment failures).

[0069] Specifically, data is sorted by timestamp to generate a time series log. For example, a user scheduled a washing machine at 2:00 AM for three consecutive days, selecting the "large wash" mode, but payment failed three times.

[0070] In detail, the dynamic time warping algorithm is an algorithm used to align time series of different lengths or speeds. It achieves the best match by calculating the minimum cumulative distance. The purpose is to eliminate differences in user operation rhythms (such as some users operate quickly and some operate slowly) and unify them for comparison with a standard template.

[0071] Specifically, direct comparison of the original time series may lead to misjudgment due to time offset (for example, user A makes an appointment at 10:00, and user B makes an appointment at 10:05, and their actual behavior patterns are the same).

[0072] In detail, define the standard operation template:

[0073] Collect historical data of N devices during the normal operation and maintenance cycle (e.g., 1,000 records);

[0074] The appointment time period can be estimated by Gaussian kernel density to generate a time distribution curve (e.g., the peak period is 18:00-20:00);

[0075] Mode sequence refers to extracting common mode switching paths (e.g., “quick wash → spin” is the optimal path);

[0076] The payment success rate needs to be statistically normally distributed (e.g., the normal payment success rate is >95%).

[0077] In an embodiment of the present invention, the method of aligning the user operation time series log with a preset standard operation template using a dynamic time warping algorithm and calculating the operation behavior deviation as the first weighting factor of the device health includes:

[0078] Extracting the reservation time period distribution density, mode switching frequency, and payment interruption times from the user operation time series log as key behavior indicators;

[0079] Calculating the cumulative path distance between each of the key behavior indicators and the preset standard operation template using a dynamic time warping algorithm;

[0080] A weighted sum is performed on the accumulated path distances to obtain an operation behavior deviation, and the operation behavior deviation is used as a first weight factor of the device health.

[0081] For example, after user A's operation sequence is aligned with the template, the cumulative distance is 5 (low deviation); user B operates in the early morning, and the cumulative distance is 20 (high deviation).

[0082] Specifically, the deviation is a quantitative value of the difference between the user operation sequence and the standard template, which is used to characterize the probability of anomaly. The purpose is to convert the timing alignment results into a comparable numerical indicator (the first weight factor F1), because directly using the DTW distance cannot reflect the comprehensive anomalies in multiple dimensions (time period, mode, payment).

[0083] In detail, the reservation time distribution density refers to the frequency of users making reservations during non-peak hours (such as early morning); the mode switching frequency refers to the frequent switching of modes in a short period of time (such as switching 3 times within 5 minutes); and the number of payment interruptions refers to the proportion of payment failures or cancellations.

[0084] Specifically, weights are set (e.g., time period distribution accounts for 50%, mode switching accounts for 30%, and payment interruption accounts for 20%), and the deviation of the operation behavior is calculated according to the following formula:

[0085]

[0086] For example, the deviation calculation of user C is:

[0087] Time deviation: 0.8 (midnight appointment)

[0088] Mode deviation: 0.6 (frequent switching)

[0089] Payout deviation: 0.9 (3 failures)

[0090] The comprehensive deviation is

[0091] In detail, the first weight factor F1 is a quantitative value of the probability of abnormal user behavior, which is used for subsequent health scoring, mapping the deviation to the health scoring system, and reflecting the weight of the impact of user behavior on device health.

[0092] In detail, user misoperation (e.g., incorrect mode selection) may be related to equipment failure (e.g., drainage blockage), and the weight needs to be adjusted dynamically.

[0093] Furthermore, the threshold rules are set as follows:

[0094] Low risk: Deviation < 0.3 → F1 = 0.1

[0095] Medium risk: 0.3 ≤ Deviation < 0.7 → F1 = 0.3

[0096] High risk: Deviation ≥ 0.7 → F1 = 0.6.

[0097] For example, user C's deviation is 0.76, which is high risk, and F1=0.6. Priority will be given to checking whether there are any abnormalities in the device he uses.

[0098] In an embodiment of the present invention, the method for generating the standard operation template includes:

[0099] Collecting user operation data of N laundry appliances during a normal operation and maintenance cycle to form a training set;

[0100] Performing Gaussian kernel density estimation on the appointment time period data in the training set to generate a time period distribution probability curve;

[0101] Extract the optimal path of washing mode switching as the standard mode sequence;

[0102] The normal distribution interval of the statistical payment success rate is used as the benchmark payment behavior indicator;

[0103] A preset standard operation template is generated based on the time period distribution probability curve, standard pattern sequence and benchmark payment behavior indicator.

[0104] Specifically, user operation data of multiple laundry appliances during normal operation and maintenance (e.g., data from 1,000 appliances for three consecutive months) is collected to establish a normal behavior baseline for comparison in subsequent anomaly detection.

[0105] In detail, the data fields are reservation timestamp, washing mode sequence, and payment status (success / failure).

[0106] For example, during a normal cycle of a certain device, the user reservation period is concentrated between 18:00 and 20:00, and the commonly used mode is "standard wash → dehydration".

[0107] In detail, Gaussian kernel density estimation generates a time period distribution curve, a non-parametric statistical method used to estimate the probability density distribution of user reservation time periods. It can identify the distribution patterns of normal time periods (such as peak hours) and detect abnormal time periods (such as high frequency of reservations in the early morning).

[0108] Specifically, a time period probability density curve (with a peak at 18:00-20:00) is generated based on all appointment timestamps (e.g., 10,000 records).

[0109] In detail, the calculation formula of Gaussian kernel density estimation is:

[0110]

[0111] in, It's on point The estimated probability density value at , is the sample size, that is, the total number of observations used for estimation (e.g., 1000 user appointment time records), is a parameter that controls the smoothness of the kernel function. is the sum index, traversing all sample points (from i=1 to i=N), is the Gaussian kernel function, is the target point (such as a specific time point) for which the probability density needs to be estimated. It is observation data points (such as user's appointment time).

[0112] In detail, each data point right The density estimate at contributes a Gaussian distribution, and the final density value is the weighted average of all contributions.

[0113] In detail, the most common paths of user mode switching (such as "quick wash → rinse → spin") are analyzed, because abnormal switching (such as frequent "quick wash → pause → quick wash") may indicate misoperation or equipment failure.

[0114] For example, in normal mode, the probability of users switching from "standard wash" to "spin" is 85%, while abnormal users frequently switch between "fast wash" and "pause".

[0115] In detail, the normal distribution interval of the statistical payment success rate is to determine the range of normal payment success rate (such as mean 95%, standard deviation 2%), because an abnormal increase in payment failure rate may reflect system failure or malicious attack.

[0116] Specifically, the mean and standard deviation of the payment success rate are calculated, and a confidence interval is set (e.g., a 95% confidence interval is 91%-99%) to obtain a normal distribution interval of the payment success rate.

[0117] Furthermore, the time period distribution curve, pattern path, and payment interval are integrated to form a multi-dimensional benchmark.

[0118] The multi-source state feature extraction module 102 is used to construct an equipment state matrix of the laundry appliance based on multi-source operation data acquired in real time, and extract characteristic parameters exceeding a preset safety threshold in the equipment state matrix to generate a second weighting factor representing an abnormal physical state of the appliance.

[0119] In an embodiment of the present invention, the process of constructing the device state matrix of the laundry device based on multi-source operation data acquired in real time includes:

[0120] Acquire multi-source operation data of the laundry device in real time, wherein the multi-source operation data includes: vibration spectrum data and motor current waveform;

[0121] Performing wavelet packet decomposition on the vibration spectrum data to extract the energy proportion of a preset frequency band as a mechanical wear feature;

[0122] After performing fast Fourier transform on the motor current waveform, the amplitude ratio of the fundamental wave to the third harmonic is calculated as the electrical abnormality feature;

[0123] The mechanical wear features and the electrical abnormality features are aligned according to timestamps to form a two-dimensional state matrix, and the two-dimensional state matrix is ​​determined to be the device state matrix of the laundry device.

[0124] In detail, the vibration spectrum refers to the vibration signal of the device collected by the acceleration sensor; the motor current waveform refers to the motor working current recorded by the current sensor.

[0125] For example, a certain sampling can obtain a vibration spectrum (0-10kHz) and a current waveform (50Hz fundamental wave).

[0126] In detail, wavelet packet decomposition to extract mechanical wear features is to decompose the vibration signal into different frequency bands and extract the energy of specific frequency bands (such as 3-5kHz), because mechanical wear (such as bearing damage) will lead to an increase in high-frequency vibration energy.

[0127] In detail, the Daubechies wavelet basis is selected for 5-layer decomposition, and the energy proportion of the 4th sub-band (3-5kHz) of the 3rd layer is calculated. For example, the energy proportion of normal equipment is 10%, which increases to 25% when worn.

[0128] Specifically, the current signal is converted into a frequency spectrum using fast Fourier transform (FFT), and the amplitude ratio of the fundamental wave (50 Hz) to the third harmonic (150 Hz) is calculated, because harmonic distortion (such as increased amplitude of the third harmonic) may cause a short circuit in the motor winding.

[0129] In detail, the amplitude ratio is For example, R<0.1 for normal equipment and R>0.3 for faulty equipment.

[0130] In detail, the construction of a two-dimensional status matrix is ​​to align the mechanical wear characteristics (E_wear) and electrical abnormality characteristics (R) by timestamp to form a matrix in order to integrate the mechanical and electrical status and comprehensively evaluate the equipment health.

[0131] In general, the user behavior template (time period, mode, payment) is combined with the device status matrix (vibration, current) to cover the dual dimensions of "human operation" and "physical status".

[0132] In detail, the preset safety thresholds are pre-set safety critical values ​​of mechanical wear and electrical abnormality characteristics, among which the mechanical wear threshold is that the energy proportion in the 3-5kHz frequency band is >15% (the average for normal equipment is 10%), and the electrical abnormality threshold is that the amplitude ratio of the fundamental wave to the third harmonic is >0.25 (the average for normal equipment is 0.1).

[0133] In detail, the abnormality of the physical status of the equipment is quantified to avoid subjective judgment bias.

[0134] Furthermore, out-of-limit characteristic parameters are extracted. For mechanical wear detection, if the 3-5kHz energy proportion in a certain sampling is 20% (exceeding the threshold of 15%), it is marked as abnormal; for electrical anomaly detection, if the current harmonic ratio is 0.3 (exceeding the threshold of 0.25), it is marked as abnormal.

[0135] In detail, the rule for generating the second weight factor (F2) is:

[0136] Single indicator exceeds the limit: F2=0.4 (medium risk).

[0137] Both indicators exceed the limit: F2=0.8 (high risk).

[0138] Continuous limit violations: accumulated over time (e.g., three consecutive limit violations, F2=1.0).

[0139] In detail, the calculation formula of the second weight factor (F2) is as follows:

[0140]

[0141] in, is the second weighting factor, is the total number of over-limit characteristic parameters, Is the over-limit characteristic parameter identifier, is the indicator weight (60% for mechanical and 40% for electrical), is the limit-out indicator function (out of limit = 1, otherwise = 0).

[0142] For example, if in a certain inspection the mechanical limit is exceeded (weight 0.6) and the electrical limit is not exceeded (weight 0.4), the second weight factor is 0.6.

[0143] The health fusion decision module 103 is used to perform a weighted decision on the first weight factor and the second weight factor to obtain a health score of the laundry device. When the health score is lower than a first threshold, device-level anomaly detection is triggered to obtain a device-level anomaly type of the laundry device.

[0144] In an embodiment of the present invention, performing a weighted decision on the first weight factor and the second weight factor to obtain a health score of the laundry appliance includes:

[0145] Dynamically calculating a weight distribution ratio of the first weight factor and the second weight factor based on an entropy method;

[0146] The health score of the laundry appliance is calculated using a weighted summation algorithm and the weight distribution ratio, wherein the weighted summation algorithm is:

[0147]

[0148] in, is the health score of the laundry appliance, is the weight distribution ratio corresponding to the first weight factor, is the weight distribution ratio corresponding to the second weight factor, is the first weighting factor, is the second weighting factor.

[0149] In detail, weights are dynamically assigned based on the amount of information (entropy value) of the data, and indicators with greater information content have higher weights.

[0150] In detail, the information entropy The calculation formula is:

[0151]

[0152] For example, when When , the information entropy of the weight factor is as follows:

[0153]

[0154]

[0155] In detail, the calculation formula of entropy redundancy is: , then the total redundancy is:

[0156] In detail, the weight distribution ratio is:

[0157]

[0158] Specifically, for a device, F1=0.6 (user behavior high risk), F2=0.8 (device status high risk), α=0.625, and β=0.375, the health score is 0.675.

[0159] In detail, the threshold rules are assumed to be: H<0.3: healthy (green); 0.3≤H<0.6: warning (yellow); H≥0.6: fault (red).

[0160] In the embodiment of the present invention, when the health score is lower than the first threshold, device-level abnormality detection is triggered, and the device-level abnormality type of the laundry device is obtained, including:

[0161] Specifically, device-level anomaly detection refers to an independent analysis of the health status of individual laundry equipment, and determines whether it is in an abnormal operating state through multi-dimensional data fusion (user behavior deviation and equipment physical status abnormality).

[0162] Specifically, the device-level abnormality type refers to a specific fault category determined based on a combination of out-of-limit characteristic parameters, including but not limited to: mechanical abnormalities, electrical abnormalities, and user operation abnormalities.

[0163] Furthermore, the mechanical anomalies include: bearing wear and belt loosening. Bearing wear refers to the proportion of high-frequency energy in the vibration spectrum exceeding a threshold (such as energy in the 3-5kHz frequency band >15%); belt loosening refers to an abnormal increase in the low-frequency vibration amplitude;

[0164] Furthermore, electrical anomalies include: motor winding short circuit and power module failure. A motor winding short circuit refers to the current harmonic ratio (third harmonic / fundamental) exceeding a threshold (e.g., R>0.25); a power module failure refers to an abnormal current waveform distortion rate.

[0165] Furthermore, user operation anomalies refer to malicious mode switching and non-time period usage. Malicious mode switching refers to frequent switching of washing modes in a short period of time (such as switching ≥3 times within 5 minutes); non-time period usage refers to high-frequency reservation of equipment in the early morning period (such as 00:00-05:00).

[0166] Example: If the health score H = 0.2 (lower than the threshold of 0.3), the motor harmonic ratio R = 0.3 (exceeding the limit), and the high-frequency vibration energy accounts for 20%, then it is determined to be a combined abnormality type of "motor winding short circuit + bearing wear".

[0167] The abnormal propagation topology modeling module 104 is used to construct a fusion topology diagram of the physical connection topology and data communication topology of the laundry equipment, dynamically allocate the abnormal propagation weight of each node in the fusion topology diagram based on the historical failure rate of the equipment and the real-time load status, and generate an abnormal propagation path probability diagram according to the connection strength between the nodes of the fusion topology diagram and the abnormal propagation weight.

[0168] In detail, the nodes of the physical connection topology are each laundry appliance, the edges are the physical connection relationships (such as shared water supply pipes and circuits), and the weights are the connection strengths (such as water pipe diameter and circuit capacity).

[0169] For example, if equipment A and equipment B share a common main water supply pipe, the connection strength is 0.8 (out of a full value of 1.0).

[0170] Specifically, the nodes of the data communication topology are laundry appliances and network devices (such as gateways and servers), the edges are data transmission paths (such as Wi-Fi and Ethernet), and the weights are network bandwidth or latency (for example, a bandwidth of 100 Mbps corresponds to a strength of 0.9).

[0171] In detail, if two nodes have both physical and data connections, they are merged into one edge, and the total connection strength is the weighted sum. For example, if the physical strength is 0.8, the data strength is 0.9, and the weight is 0.6, the total connection strength is 0.84.

[0172] In detail, the propagation probability formula is , where is the propagation weight of target node j, is a node arrive The connection strength, is a node arrive The probability of transmission.

[0173] For example, if device A (weight 0.6) is connected to device B (weight 0.8) with a connection strength of 0.72, the transmission probability is 0.576.

[0174] In an embodiment of the present invention, the method of dynamically allocating the abnormality propagation weight of each node in the fusion topology graph based on the historical failure rate of the device and the real-time load status includes:

[0175] Determine the initial static weight of each node according to the betweenness centrality and historical failure rate of the laundry device in the fusion topology graph;

[0176] When the real-time load state of the laundry appliance exceeds 70% of the rated value, determining a dynamic correction coefficient of the initial static weight, and correcting the node weights in the fusion topology graph based on the dynamic correction coefficient;

[0177] When the laundry appliance is in a maintenance lock state, the node weights of the fusion topology graph are forcibly set to zero to block abnormal propagation.

[0178] In detail, betweenness centrality measures the hubness of a node in the graph, and the calculation formula is as follows:

[0179]

[0180] in, is the betweenness centrality, is a node To Node The number of shortest paths, Yes, it passes through the node The number of shortest paths, It is a node.

[0181] Specifically, the historical failure rate is the ratio of the number of past failures of the equipment to the total operating time.

[0182] Specifically, the trigger condition for real-time load correction is that the real-time load of the device is greater than 70% of the rated value.

[0183] For example, if the real-time load of device B is 85% and the initial weight is 0.4, the dynamic correction coefficient is:

[0184]

[0185] Furthermore, .

[0186] Specifically, if a device is locked for maintenance, its propagation weight is forced to 0. For example, if device C is locked for maintenance, the weight changes from 0.5 to 0, and the anomaly cannot be propagated through C.

[0187] The spatiotemporal warning visualization module 105 is used to calculate the group abnormality probability of the laundry equipment based on the equipment-level abnormality type and the abnormal propagation path probability map, and divide the multi-level alarm level according to the probability interval, and integrate the location information of the laundry equipment with the multi-level alarm level to generate a visual alarm log.

[0188] In an embodiment of the present invention, the calculation of the group abnormality probability of the laundry devices based on the device-level abnormality type and the abnormality propagation path probability graph includes:

[0189] The device-level anomaly type and the anomaly propagation path probability map are input into a spatiotemporal propagation model to calculate the group anomaly probability of the laundry devices, wherein the spatiotemporal propagation model satisfies:

[0190] For mechanical anomalies, an exponential decay propagation rule based on physical distance is adopted;

[0191] For data anomalies, a linear propagation rule based on the number of network hops is adopted;

[0192] When two types of anomalies coexist, mechanical anomaly propagation calculation is performed first and the data anomaly correction term is superimposed.

[0193] In detail, the decay factor used in the exponential decay propagation rule may be a time decay factor or a distance decay factor.

[0194] In detail, the exponential decay propagation rule based on physical distance includes: physical distance decay factor. The calculation formula is as follows:

[0195]

[0196] in, is the initial propagation probability (determined by device weight and connection strength), is the distance attenuation coefficient (such as 0.1 / m), is the physical distance between devices (in meters).

[0197] Specifically, the linear propagation rules of data-type anomalies include: network hop number decay, where network hop number decay The calculation formula is as follows:

[0198]

[0199] in, is the network hop count attenuation, is the initial propagation probability (determined by device weight and connection strength), is the decay rate per hop (e.g. 0.15 / hop), is the number of network hops (the length of the path from the source device to the destination device).

[0200] In detail, the priority mechanical anomaly propagation refers to the superposition of a data anomaly correction term based on the mechanical anomaly propagation probability PmechPmech.

[0201] In this embodiment of the present invention, the calculation formula for the group abnormality probability is as follows:

[0202]

[0203] in, is the population abnormality probability of the laundry equipment, It is due to the initial abnormal device the number of devices that may fail due to the spread of is the number of all laundry appliances in a specific area, is the equipment number of the affected laundry equipment, It's laundry equipment The exception propagates to the laundry equipment The final probability of is the equipment number of the laundry equipment in a specific area, It's laundry equipment the probability of an anomaly occurring independently (independent of propagation), It's laundry equipment. is the device number of the device that initially failed.

[0204] Specifically, the formula for calculating the group anomaly probability consists of two main parts: the probability of cascading failures and the probability of concurrent regional anomalies. The total group anomaly probability is then calculated using the product-and-complement method.

[0205] In detail, is the group abnormality probability of the laundry appliances, indicating the possibility of an abnormality occurring in the entire laundry appliance group. The closer the value is to 1, the higher the abnormality risk; It is due to the initial abnormal device The number of devices that may fail due to the spread of The exception propagates to 3 devices, then , is the number of laundry facilities in a specific area (e.g. a dormitory building), is the equipment number of the affected laundry equipment. , indicating the second affected device.

[0206] In detail, the cascading failure probability Indicates that all affected devices are not affected by the device The probability of failure due to abnormality; the probability of regional concurrent abnormality The sum of the probabilities of independent device anomalies in the area is set to 1 if it exceeds 1 (probability upper limit constraint).

[0207] For example, suppose there are two devices in the system, where the probability of propagating anomalies is ,

[0208] Independent anomaly probability ,

[0209] but .

[0210] In detail, express An independent individual or local abnormality judgment unit ( is the final abnormal probability of a single unit) and the joint probability that no abnormality occurs; express Correction items is the abnormal probability corresponding to the correction term) and the joint probability that no abnormality occurs.

[0211] In an embodiment of the present invention, the step of fusing the location information of the laundry appliance with the multi-level alarm levels to generate a visual alarm log includes:

[0212] Rendering a three-level thermal map of abnormal laundry equipment on a campus electronic map based on the location information of the laundry equipment and the multi-level alarm level, where red, yellow, and blue correspond to high risk, medium risk, and low risk, respectively;

[0213] Mark the possible impact range and estimated arrival time of the abnormal propagation of the abnormal device;

[0214] Automatically generate maintenance work orders containing spare parts replacement lists and operation instruction videos based on the pre-associated maintenance knowledge base.

[0215] In detail, based on the location information of laundry equipment and combined with multi-level alarm levels (high, medium, and low risk), a thermal layer of abnormal equipment is rendered with color differentiation on the campus electronic map, where red represents high risk, yellow represents medium risk, and blue represents low risk; through the visual heat map, the distribution density and risk level of abnormal laundry equipment are intuitively displayed, which facilitates the rapid positioning of high-risk areas and assists management decisions.

[0216] In detail, the possible impact range of abnormal spread of abnormal equipment is marked on the electronic map, and the estimated arrival time is predicted; early warning of abnormal spread trends is given to help managers plan resource allocation, such as guiding students to avoid affected areas in advance or making emergency preparations. The device location information is GPS coordinates or building number + room number.

[0217] Specifically, use GIS platforms (such as ArcGIS, Mapbox) or open source libraries (such as Leaflet.js) to render maps; integrate real-time data interfaces (such as MQTT or WebSocket) to dynamically update device status.

[0218] In detail, a gradient color coverage area is generated according to the device density and the alarm level.

[0219] Example: A dormitory building has densely populated equipment and many high-risk devices, which are displayed as dark red blocks; scattered low-risk devices are displayed as light blue dots.

[0220] Specifically, it links with a pre-established maintenance knowledge base to automatically generate maintenance work orders containing spare parts replacement lists and operation guidance videos; with standardized maintenance processes, maintenance personnel can directly obtain spare parts information and operation instructions based on the work order, reducing maintenance preparation time, improving fault handling efficiency, and reducing dependence on maintenance personnel's experience.

[0221] Specifically, the input triggers for the work order generation logic are the equipment alarm level and anomaly type (such as "mechanical wear"). The matching rule is: if the anomaly type is "motor harmonic ratio exceeds the standard", match the "motor failure" entry in the knowledge base, and extract the spare parts list (such as "motor winding") and video link.

[0222] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0223] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to achieve optimal results.

[0224] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A campus self-service laundry equipment abnormality identification system based on the Internet of Things, characterized by: The system comprises: A user operation behavior alignment module is used to generate a user operation time series log of a laundry device, align the user operation time series log with a preset standard operation template through a dynamic time warping algorithm, calculate the operation behavior deviation and generate a first weight factor representing the probability of abnormal user behavior, wherein the alignment of the user operation time series log with the preset standard operation template through the dynamic time warping algorithm and the calculation of the operation behavior deviation as the first weight factor of the device health include: extracting the reservation time period distribution density, mode switching frequency and payment interruption number in the user operation time series log as key behavior indicators; calculating the cumulative path distance between each of the key behavior indicators and the preset standard operation template through the dynamic time warping algorithm; performing weighted summation on the cumulative path distance to obtain the operation behavior deviation, and using the operation behavior deviation as the first weight factor of the device health; a multi-source state feature extraction module, configured to construct a device state matrix of the laundry appliance based on multi-source operation data acquired in real time, and extract characteristic parameters exceeding a preset safety threshold in the device state matrix to generate a second weighting factor representing an abnormal physical state of the appliance; a health fusion decision module, configured to perform a weighted decision on the first weight factor and the second weight factor to obtain a health score of the laundry appliance, and trigger device-level anomaly detection when the health score is lower than a first threshold to obtain a device-level anomaly type of the laundry appliance; An anomaly propagation topology modeling module is used to construct a fusion topology diagram of the physical connection topology and data communication topology of the laundry device, dynamically assign an anomaly propagation weight to each node in the fusion topology diagram based on the historical failure rate of the device and the real-time load status, and generate an anomaly propagation path probability diagram based on the connection strength between the nodes in the fusion topology diagram and the anomaly propagation weight; The spatiotemporal warning visualization module is used to calculate the group anomaly probability of the laundry equipment based on the equipment-level anomaly type and the anomaly propagation path probability map, divide the alarm level into multiple levels according to the probability interval, and integrate the location information of the laundry equipment with the multiple levels of alarm levels to generate a visual alarm log.

2. The campus self-service laundry equipment abnormality identification system based on the Internet of Things according to claim 1 is characterized in that: The user operation behavior alignment module generates a user operation time sequence log of the laundry device, including: Collecting user operation behavior data, wherein the user operation behavior data includes: reservation time period, washing mode selection and payment record; A user operation time series log including a reservation period, a washing mode selection sequence and a payment success rate is generated based on the user operation behavior data.

3. The campus self-service laundry equipment abnormality identification system based on the Internet of Things as claimed in claim 1, characterized in that: The method for generating a preset standard operation template in the user operation behavior alignment module includes: Collecting user operation data of N laundry appliances during a normal operation and maintenance cycle to form a training set; Performing Gaussian kernel density estimation on the appointment time period data in the training set to generate a time period distribution probability curve; Extract the optimal path of washing mode switching as the standard mode sequence; The normal distribution interval of the statistical payment success rate is used as the benchmark payment behavior indicator; A preset standard operation template is generated based on the time period distribution probability curve, standard pattern sequence and benchmark payment behavior indicator.

4. The campus self-service laundry equipment abnormality identification system based on the Internet of Things as claimed in claim 1, characterized in that: When the multi-source state feature extraction module constructs the device state matrix of the laundry device based on the multi-source operation data acquired in real time, the multi-source state feature extraction module includes: Acquire multi-source operation data of the laundry device in real time, wherein the multi-source operation data includes: vibration spectrum data and motor current waveform; Performing wavelet packet decomposition on the vibration spectrum data to extract the energy proportion of a preset frequency band as a mechanical wear feature; After performing fast Fourier transform on the motor current waveform, the amplitude ratio of the fundamental wave to the third harmonic is calculated as the electrical abnormality feature; The mechanical wear features and the electrical abnormality features are aligned according to timestamps to form a two-dimensional state matrix, and the two-dimensional state matrix is ​​determined to be the device state matrix of the laundry device.

5. The campus self-service laundry equipment abnormality identification system based on the Internet of Things as claimed in claim 1, characterized in that: When the health fusion decision module performs a weighted decision on the first weight factor and the second weight factor to obtain the health score of the laundry appliance, the module includes: Dynamically calculating a weight distribution ratio of the first weight factor and the second weight factor based on an entropy method; The health score of the laundry appliance is calculated using a weighted summation algorithm and the weight distribution ratio, wherein the weighted summation algorithm is: in, is the health score of the laundry appliance, is the weight distribution ratio corresponding to the first weight factor, is the weight distribution ratio corresponding to the second weight factor, is the first weighting factor, is the second weighting factor.

6. The campus self-service laundry equipment abnormality identification system based on the Internet of Things according to claim 1, characterized in that: The abnormal propagation topology modeling module, when dynamically allocating abnormal propagation weights of each node in the fusion topology graph based on the historical failure rate of the equipment and the real-time load status, includes: Determine the initial static weight of each node according to the betweenness centrality and historical failure rate of the laundry device in the fusion topology graph; When the real-time load state of the laundry appliance exceeds 70% of the rated value, determining a dynamic correction coefficient of the initial static weight, and correcting the node weights in the fusion topology graph based on the dynamic correction coefficient; When the laundry appliance is in a maintenance lock state, the node weights of the fusion topology graph are forcibly set to zero to block abnormal propagation.

7. The campus self-service laundry equipment abnormality identification system based on the Internet of Things as claimed in claim 1, characterized in that: When the spatiotemporal warning visualization module calculates the group abnormality probability of the laundry devices based on the device-level abnormality type and the abnormal propagation path probability graph, the module includes: The device-level anomaly type and the anomaly propagation path probability map are input into a spatiotemporal propagation model to calculate the group anomaly probability of the laundry devices, wherein the spatiotemporal propagation model satisfies: For mechanical anomalies, an exponential decay propagation rule based on physical distance is adopted; For data anomalies, a linear propagation rule based on the number of network hops is adopted; When two types of anomalies coexist, mechanical anomaly propagation calculation is performed first and the data anomaly correction term is superimposed.

8. The campus self-service laundry equipment abnormality identification system based on the Internet of Things according to claim 7, characterized in that: The calculation formula of the group abnormality probability in the spatiotemporal warning visualization module is as follows: in, is the population abnormality probability of the laundry equipment, Is due to initial abnormal equipment the number of devices that may fail due to the spread of is the number of all laundry appliances in a specific area, is the equipment number of the affected laundry equipment, It's laundry equipment The exception propagates to the laundry equipment The final probability of is the equipment number of the laundry equipment in a specific area, It's laundry equipment The probability of an independent anomaly occurring, It's laundry equipment. is the device number of the device that initially failed.

9. A campus self-service laundry equipment abnormality identification system based on the Internet of Things according to any one of claims 1 to 7, characterized in that: When the spatiotemporal warning visualization module generates a visual alarm log by fusing the location information of the laundry appliance with the multi-level alarm levels, the module includes: Rendering a three-level thermal map of abnormal laundry equipment on a campus electronic map based on the location information of the laundry equipment and the multi-level alarm level, where red, yellow, and blue correspond to high risk, medium risk, and low risk, respectively; Mark the possible impact range and estimated arrival time of the abnormal propagation of the abnormal device; Automatically generate maintenance work orders containing spare parts replacement lists and operation instruction videos based on the pre-associated maintenance knowledge base.

Images