Static electricity detection method, system and equipment for cableway steel cable and storage medium

Abstract

The invention relates to the technical field of data processing, and particularly provides an electrostatic detection method, system and device for a cableway steel cable and a storage medium, and the method comprises the steps: collecting the multi-modal data of the cableway steel cable through a distributed sensor network under the working condition of no electrostatic risk; based on the multi-modal data, training a dynamic reference model; in the cableway operation process, real-time multi-mode data are synchronously collected, and a corresponding target normal charge quantity range is obtained from the dynamic reference model according to the real-time multi-mode data; carrying out self-adaptive noise reduction processing on a real-time mixed electrostatic signal in the real-time multi-modal data so as to extract an electrostatic signal after noise reduction; and comparing the electrostatic charge quantity corresponding to the electrostatic signal with the target normal charge quantity range, and determining a risk level based on a comparison result. According to the invention, through construction of the dynamic reference model and intelligent signal processing, static detection of the cableway steel cable is realized.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, specifically relating to a method, system, equipment, and storage medium for electrostatic detection of cableway steel cables. Background Technology

[0002] During operation, cableway cables accumulate static charge due to friction and stripping, posing significant safety risks such as fire and electric shock. Existing electrostatic discharge (ESD) detection methods largely rely on single-point ESD voltage measurement and fixed threshold alarms, which have significant limitations in practical applications. Firstly, complex mechanical vibrations and electromagnetic interference in industrial environments severely contaminate ESD signals, leading to inaccurate measurements and high false alarm rates. Secondly, the ESD characteristics of the cable are dynamically affected by multiple operating conditions such as operating speed, ambient temperature, and humidity; fixed thresholds cannot adapt adaptively, easily resulting in missed alarms or untimely warnings. Furthermore, traditional methods lack the ability to assess charge accumulation trends and actively neutralize them, remaining largely in a passive alarm state. Therefore, there is an urgent need for an advanced detection method that can overcome environmental interference, adapt to changing operating conditions, and achieve proactive safety protection. Summary of the Invention

[0003] In view of the above-mentioned shortcomings of the prior art, the present invention provides a method, system, equipment and storage medium for electrostatic detection of cableway steel cables to solve the above-mentioned technical problems.

[0004] In a first aspect, the present invention provides a method for electrostatic detection of cableway steel cables, comprising: Under conditions without electrostatic risks, multimodal data of the cableway steel cable is collected through a distributed sensor network; based on the multimodal data, a dynamic benchmark model that can characterize the mapping relationship between the real-time operating state and the normal charge range is trained. During cableway operation, real-time multimodal data is collected synchronously, and the corresponding target normal charge range is obtained from the dynamic reference model based on the real-time multimodal data. Adaptive noise reduction processing is performed on the real-time mixed electrostatic signals in the real-time multimodal data to extract the noise-reduced electrostatic signals; The electrostatic charge corresponding to the electrostatic signal is compared with the target normal charge range, and the risk level is determined based on the comparison result.

[0005] In an optional implementation, under conditions without electrostatic risk, multimodal data of the cableway steel cable is collected via a distributed sensor network, including: Select at least one of the following operating conditions: high ambient humidity, rainy weather, or low-speed unloaded cableway operation. Continuously collect data on the electrostatic signal, vibration signal, electromagnetic interference signal, ambient temperature and humidity, wind speed, and cableway operating speed for 24 to 72 hours.

[0006] In an optional implementation, a dynamic benchmark model is trained based on the multimodal data. This dynamic benchmark model can characterize the mapping relationship between operating parameters and the normal charge range, including: Construct a spatiotemporal graph of sensors, where nodes are the various sensor measurement points in a distributed sensor network, node features are multimodal data based on time-series windows, and edges are defined according to the physical location relationship or signal correlation between sensors; The spatiotemporal graph is input into the spatiotemporal graph attention network. Through the graph attention mechanism in the spatial dimension and the sequence model in the temporal dimension, the low-dimensional spatiotemporal embedding features of each node are learned and output. Based on the low-dimensional spatiotemporal embedding features of all nodes, a deep embedding clustering algorithm is used for unsupervised clustering, where the clustering loss and network reconstruction loss are jointly optimized. Based on the clustering results, the normal charge range corresponding to each cluster is calculated to generate the dynamic baseline model.

[0007] In an optional implementation, obtaining the corresponding target normal charge range from the dynamic reference model based on the real-time multimodal data includes: The real-time spatiotemporal graph constructed based on real-time multimodal data is input into a trained spatiotemporal graph attention network encoder to obtain real-time low-dimensional spatiotemporal embedding features. Calculate the similarity between the real-time embedded feature and the cluster centers of each cluster in the dynamic benchmark model; The range of normal charge corresponding to the cluster with the highest similarity is taken as the target range of normal charge corresponding to the current real-time operating condition.

[0008] In an optional implementation, adaptive noise reduction processing is performed on the real-time mixed electrostatic signals in the real-time multimodal data to extract the noise-reduced electrostatic signals, including: Vibration signals and electromagnetic interference signals are extracted from the real-time multimodal data; The vibration signal and electromagnetic interference signal are used as reference noise inputs. The signal components related to the reference noise are canceled out in real time from the real-time mixed electrostatic signal through a normalized minimum mean square adaptive filter, so as to output a noise-reduced electrostatic signal.

[0009] In an optional implementation, the electrostatic charge corresponding to the electrostatic signal is compared with the target normal charge range, and the risk is determined based on the comparison result, including: Calculate the real-time electrostatic charge based on the denoised electrostatic signal; The real-time electrostatic charge quantity is compared with the current normal electrostatic charge quantity range obtained from the dynamic reference model; Based on the comparison results, the risk level is determined according to the following rules: If the real-time electrostatic charge is within the normal range, it is determined to be in a normal state; If the real-time electrostatic charge exceeds the normal range but is below the first threshold, or the charge accumulation rate exceeds the rate threshold, it is determined to be in a state of alert. If the real-time electrostatic charge reaches or exceeds the first threshold, or if an intermittent discharge pulse is detected, it is determined to be a warning state and an audible and visual warning is triggered. If the real-time electrostatic charge reaches or exceeds a second threshold higher than the first threshold, or if continuous strong discharge is detected, it is determined to be an alarm state and the highest level alarm is triggered.

[0010] In one optional implementation, the real-time electrostatic charge is calculated based on the denoised electrostatic signal, including: For the electrostatic signal after adaptive noise reduction processing, the effective value of its voltage amplitude is calculated according to a preset time window, and the effective value is used as the real-time electrostatic potential. Based on the real-time electrostatic potential, and combined with the dynamic capacitance value of the steel cable to the ground that is pre-established through simulation or experiment and matches the current working conditions, the real-time electrostatic charge is calculated using the formula Q=C×V; where Q is the charge, C is the dynamic capacitance value, and V is the real-time electrostatic potential.

[0011] Secondly, the present invention provides an electrostatic detection system for cableway steel cables, comprising: The benchmark construction module is used to collect multimodal data of the cableway steel cable through a distributed sensor network under conditions without electrostatic risks; based on the multimodal data, a dynamic benchmark model that can characterize the mapping relationship between the real-time operating state and the normal charge range is trained. The real-time detection module is used to synchronously collect real-time multimodal data during cableway operation, and obtain the corresponding target normal charge range from the dynamic reference model based on the real-time multimodal data. The signal denoising module is used to perform adaptive denoising processing on the real-time mixed electrostatic signals in the real-time multimodal data in order to extract the denoised electrostatic signals. The risk determination module is used to compare the electrostatic charge corresponding to the electrostatic signal with the target normal charge range, and determine the risk level based on the comparison result.

[0012] Thirdly, a device is provided, comprising: A memory used to store the electrostatic detection program for cableway steel cables; A processor is configured to implement the steps of the electrostatic detection method for the cableway as provided in the first aspect when executing the electrostatic detection program for the cableway.

[0013] Fourthly, a computer-readable storage medium is provided, on which an electrostatic detection program for a cableway is stored, wherein when the electrostatic detection program for the cableway is executed by a processor, the steps of the electrostatic detection method for the cableway as provided in the first aspect are implemented.

[0014] The beneficial effects of this invention are as follows: the electrostatic detection method, system, equipment, and storage medium for cableway steel cables provided by this invention achieve multiple beneficial effects in electrostatic detection of cableway steel cables by constructing a dynamic benchmark model and intelligent signal processing. First, the use of multi-source signal fusion and adaptive noise reduction technology effectively suppresses vibration and electromagnetic interference, significantly improving signal quality and system anti-interference capability. Second, based on deep learning-based state recognition and dynamic benchmarks, the system can adapt to changes in operating conditions, fundamentally avoiding false alarms and missed alarms caused by fixed thresholds. Furthermore, through charge quantization and multi-level risk assessment, a shift from post-event alarm to pre-event warning is achieved. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic flowchart of a method according to an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram illustrating an application scenario of a method according to an embodiment of the present invention.

[0018] Figure 3 This is a schematic block diagram of a system according to an embodiment of the present invention.

[0019] Figure 4 This is a schematic diagram of the structure of a device provided in an embodiment of the present invention.

[0020] Among them, 1. Cableway support; 2. Cable wheel primary shaft; 3. Electrostatic detection sensor. Detailed Implementation

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

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0023] The electrostatic detection method for cableway steel cables provided in this embodiment of the invention is executed by computer equipment, and correspondingly, the electrostatic detection system for cableway steel cables runs in the computer equipment.

[0024] Figure 1 This is a schematic flowchart illustrating a method according to an embodiment of the present invention. Wherein, Figure 1 The implementing entity can be an electrostatic detection system for cableway steel cables. Depending on different requirements, the order of steps in this flowchart can be changed, and some steps can be omitted.

[0025] like Figure 1 As shown, the method includes: S1. Under conditions without electrostatic risks, multimodal data of the cableway steel cable is collected through a distributed sensor network; based on the multimodal data, a dynamic benchmark model that can characterize the mapping relationship between the real-time operating state and the normal charge range is trained. S2. During the operation of the cableway, real-time multimodal data is collected synchronously, and the corresponding target normal charge range is obtained from the dynamic reference model based on the real-time multimodal data; S3. Perform adaptive noise reduction processing on the real-time mixed electrostatic signal in the real-time multimodal data to extract the noise-reduced electrostatic signal; S4. Compare the electrostatic charge corresponding to the electrostatic signal with the target normal charge range, and determine the risk level based on the comparison result.

[0026] In one embodiment of the present invention, based on step S1, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0027] S101. Collect multimodal data of the cableway steel cable through a distributed sensor network, including: Operating Condition Selection: To ensure that the collected data accurately reflects the "normal state" without static electricity buildup, one or more of the following operating conditions should be selected first: High ambient humidity weather, such as when the relative humidity of the air is consistently above 80%.

[0028] On rainy days, natural precipitation is used to keep the surface of the steel cable moist, thereby greatly inhibiting the generation and accumulation of static charge.

[0029] When the cableway is running at low speed and unloaded, the intensity of triboelectric charging is minimized by reducing the operating speed and removing the load.

[0030] Data collection and recording: Initiate the deployment of a distributed sensor network at key locations along the cableway (such as drive stations, detour stations, and support structures). (Reference) Figure 2 The cableway consists of a support frame 1 and steel cables. Electrostatic detection sensors 3 are deployed on the steel cables at key points that contact the primary shaft 2 of the cable pulley.

[0031] The network synchronously collects the following multimodal data: electrostatic signals measured by non-contact electrostatic sensors, vibration signals measured by vibration accelerometers, electromagnetic interference signals measured by EMI probes, environmental parameters measured by temperature and humidity sensors and anemometers, and cableway operating speed provided by encoders.

[0032] Data should be collected continuously for 24 to 72 hours. This duration is designed to cover a complete diurnal cycle and possible fluctuations in environmental conditions, ensuring the comprehensiveness and statistical significance of the acquired dataset. The acquisition frequency can be set as needed, for example, 1 kHz, to capture sufficient dynamic detail.

[0033] After initial processing and analog-to-digital conversion at each data acquisition node, all data is transmitted via industrial Ethernet to a central processing server for storage, preparing for subsequent training of the dynamic benchmark model.

[0034] S102. Based on the multimodal data, a dynamic benchmark model is trained. This dynamic benchmark model can characterize the mapping relationship between operating parameters and the normal charge range, including: 1. Construction of the spatiotemporal graph structure A spatiotemporal graph of sensors is constructed based on the collected multimodal data. Specifically, each sensor measurement point in the distributed sensor network (such as the electrostatic sensor on support 1, the vibration sensor on support 2, etc.) is defined as a node in the graph. The feature vector of each node is composed of its corresponding multimodal time-series data: the observation values ​​of each sensor at continuous time steps are taken to form a feature vector based on a preset time window (such as 60 seconds), which includes electrostatic, vibration, EMI signals and environmental parameters within that time period. The edges between nodes are defined according to one or a combination of the following two methods: one is based on the physical positional relationship between sensors, connecting sensors on adjacent supports to each other; the other is based on signal correlation, calculating the Pearson correlation coefficient between signals from different sensors, and if the coefficient exceeds a predetermined threshold (such as 0.7), then a connection edge is established between the corresponding nodes. Thus, a graph structure G=(V,E) representing the physical topology and information flow of the cableway is constructed, where V is the set of nodes and E is the set of edges.

[0035] 2. Deep Spatiotemporal Feature Learning The constructed spatiotemporal graph is input into a spatiotemporal graph attention network for encoding. Within the network: Spatial Dimension: A graph attention mechanism is adopted, where each node calculates its attention weight α with its neighboring nodes. ij Using the LeakyReLU activation function, weighted aggregation of neighbor features is used to model spatial dependencies. The formula can be expressed as: in is the feature after aggregation of node i, N(i) is its neighbor set, W is the weight matrix, and σ is the non-linear activation function; This refers to the feature vector of node i's neighbor node j.

[0036] Temporal dimension: The aggregated feature sequence of each node is input into a gated recurrent unit (optional temporal convolutional network) to capture its evolution pattern over time. GRU learns long-term dependencies through its gating mechanism (update gate and reset gate), and its hidden state serves as the temporal context feature of the node.

[0037] After multi-layer STGAT encoding, each node is mapped to a low-dimensional spatiotemporal embedding feature vector z that condenses spatiotemporal context information. i Its dimensionality is much lower than that of the original features, but it retains key discriminative information.

[0038] 3. Deep Embedding Clustering Optimization Based on the low-dimensional spatiotemporal embedding feature Z={z} of all nodes i Unsupervised clustering is performed using a deep embedding clustering algorithm: Initialization: First, the K-Means algorithm is used to perform an initial partitioning of Z to obtain cluster centers. .

[0039] Soft assignment: Calculate each embedded feature z i With each cluster center μ j The similarity is calculated using Student's t-distribution as the kernel function to obtain the soft-assigned probability distribution q. ij :

[0040] Target distribution and joint training: by Calculate the target distribution This distribution reinforces high-confidence assignments. The clustering loss is defined as the KL divergence: This loss is compared with the reconstruction loss L of the STGAT encoder. r (For example, Mean Squared Error) is combined according to weight λ to form the total loss L=L r +λL cBy jointly optimizing the entire network parameters and cluster centers through backpropagation, the learned features are both suitable for clustering and can retain key data information.

[0041] 4. Dynamic benchmark library generation After clustering convergence, for each final cluster Ck: Calculate the statistical range of electrostatic charge (obtained by Q=C×V) for all samples within the cluster, for example, taking the 5th and 95th percentiles as the normal charge range. .

[0042] It can also calculate and store typical characteristics of the vibration signal within the cluster, such as the average dominant frequency and background noise level.

[0043] All clusters and their corresponding normal feature ranges are stored, thus generating the dynamic benchmark model (or dynamic benchmark library). In practical applications, by analyzing the nodes and features that contribute the most to clustering, interpretability analysis of the runtime state can also be provided.

[0044] In one embodiment of the present invention, based on step S2, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0045] S201. Data Synchronization Acquisition and Spatiotemporal Map Construction.

[0046] The central processing system instructs all distributed sensor nodes to synchronously acquire real-time multimodal data at the same frequency (e.g., 1 kHz) as during the training phase. This data includes real-time mixed electrostatic signals, vibration signals, electromagnetic interference signals, and real-time ambient temperature, humidity, wind speed, and cableway operating speed at each measurement point. The system then constructs a real-time spatiotemporal graph from the multimodal time-series data of the current moment and the previous time window (e.g., 60 seconds) using a predefined graph structure (e.g., an adjacency matrix defined during initialization based on physical location or signal correlation). In this graph, nodes correspond to individual sensors, node features are their time-series data vectors, and edges represent the relationships between sensors.

[0047] S202. Spatiotemporal feature encoding and embedding representation generation.

[0048] The real-time spatiotemporal graph constructed in step S201 is input into the spatiotemporal graph attention network encoder that was pre-trained during the initialization phase. This encoder operates as follows: Spatial attention aggregation: For each node in the graph, STGAT calculates its association weight with all neighboring nodes through an attention mechanism, and aggregates the information of neighboring nodes accordingly to capture the spatial dependencies of the sensor network.

[0049] Temporal feature extraction: The aggregated feature sequence of each node is input into the gated recurrent unit to which it is attached, learns its own evolution pattern over time, and captures temporal dependencies.

[0050] After multi-layer STGAT processing, the final output is a fixed-dimensional, real-time, low-dimensional spatiotemporal embedding feature vector that condenses the spatiotemporal state information of the entire sensor network.

[0051] S203. Dynamic benchmark matching and target range acquisition.

[0052] The system calls the set of cluster centers obtained during the initialization phase of training the deep embedding clustering model. Subsequently, it calculates the cosine similarity between the real-time embedding features obtained in step S202 and each cluster center. The system selects the cluster with the highest similarity, representing the "historical normal operating condition mode" that best matches the current real-time operating status of the cableway. Finally, the system retrieves the preset normal electrostatic charge range (e.g., the 5% to 95th percentile of the charge of all samples within the cluster) from the dynamic benchmark library and uses this range as the target normal charge range for comparison and risk assessment.

[0053] In one embodiment of the present invention, based on step S3, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0054] S301. Reference noise signal extraction The system extracts vibration signals and electromagnetic interference signals as reference noise sources from synchronously acquired real-time multimodal data streams. Specifically: The vibration signal is provided by a vibration accelerometer installed next to the electrostatic sensor, which directly reflects the interference caused by the mechanical vibration of the steel cable.

[0055] Electromagnetic interference signals are provided by EMI probes deployed in areas with strong electromagnetic interference (such as near drive motors), and these signals reflect the background electromagnetic noise.

[0056] These two signals were acquired synchronously with the mixed electrostatic signal, ensuring time alignment and providing an accurate reference input for subsequent adaptive filtering.

[0057] S302. Normalized Least Mean Square Adaptive Filtering for Noise Reduction The extracted vibration and electromagnetic interference signals are used as reference noise inputs, and together with the real-time mixed electrostatic signal, are input into a normalized minimum mean square adaptive filter. The specific workflow of this filter is as follows: Filter initialization: Set the parameters of the adaptive filter, including the filter order (e.g., 64th order) and the step size parameter (e.g., μ=0.01).

[0058] Adaptive process: at each sampling time k: The filter uses two reference noise input signals x1(k) (vibration) and x2(k) (EMI) and their historical values ​​to form the reference input vector x(k).

[0059] By continuously adjusting its internal weight vector w(k), the filter output y(k) (i.e. the best estimate of the noise) is made closest to the noise component in the mixed electrostatic signal d(k).

[0060] The core weight update formula of the NLMS algorithm is:

[0061] Where, e(k) = d(k) y(k) is the error signal. Step size factor It is a small positive number used to avoid division by zero errors.

[0062] Signal output: The error signal e(k) is taken as the final output of the system, which is the denoised electrostatic signal. Physically speaking, e(k) represents the signal remaining after the noise components related to vibration and electromagnetic interference are canceled out in real time from the original mixed signal d(k), and can be considered as the real electrostatic signal component.

[0063] In one embodiment of the present invention, based on step S4, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0064] S401. Calculate the real-time electrostatic charge based on the noise-reduced electrostatic signal; For the electrostatic signal after adaptive noise reduction processing, the effective value of its voltage amplitude is calculated according to a preset time window (e.g., 1 second), and this effective value is used as the real-time electrostatic potential.

[0065] Where, x k Let be the voltage value of the denoised electrostatic signal at the k-th sampling point, and N be the number of sampling points within the time window. This effective value calculation can stably characterize the average level of the electrostatic potential.

[0066] Based on the real-time electrostatic potential, and combined with the dynamic capacitance model of the steel cable to the ground established in advance through finite element simulation or actual measurement, the dynamic capacitance value C that matches the current operating speed, cable sag and other working conditions is obtained. The real-time electrostatic charge is calculated by the relationship Q=C×V, where Q is the charge, C is the dynamic capacitance value and V is the real-time electrostatic potential.

[0067] S402. Compare the real-time electrostatic charge quantity with the current normal electrostatic charge quantity range obtained from the dynamic reference model.

[0068] The central processing system matches the corresponding target cluster from the dynamic benchmark model based on the current real-time data, and reads the normal electrostatic charge range corresponding to that cluster. Subsequently, the real-time charge quantity Q calculated by S401 is compared with this normal range.

[0069] S403. Based on the comparison results, the risk level shall be determined according to the following rules: Normal state: When At that time, the system determined it to be safe, and the interface displayed that it was running normally.

[0070] Note the following conditions: If Q exceeds the normal range but is below the preset first threshold T1, or if the system detects that the rate of change of charge within a short period of time (e.g., 30 seconds) (i.e., the charge accumulation rate ΔtΔQ) exceeds the set rate threshold R. th If the error occurs, a "Caution" sign will be prominently displayed on the monitoring interface to alert maintenance personnel.

[0071] Warning status: If Q≥T1, or the signal analysis module identifies intermittent discharge pulses in the noise-reduced electrostatic signal (characterized by a sudden increase in amplitude, duration <10ms and non-continuous occurrence), the system immediately triggers the audible and visual alarm installed in the control room to issue a yellow warning signal.

[0072] Alarm status: If Q≥T2= (where T2>T1 is a higher second threshold), or if a continuous strong discharge is detected (characterized by the continuous appearance of high-frequency pulse trains), the system triggers the highest level alarm (red warning, rapid beeping) and automatically starts the subsequent active neutralization process.

[0073] Among them, risk thresholds T1 and T2 are preset, global, absolute risk thresholds. They do not change with the operating conditions and are hard thresholds set based on safety standards. T1 is the low-risk threshold (early warning line). T2 is the high-risk threshold (alarm line), and T2 > T1.

[0074] In some embodiments, the electrostatic detection system for the cableway steel cable may include multiple functional modules composed of computer program segments. The computer programs for each program segment in the electrostatic detection system for the cableway steel cable may be stored in the memory of a computer device and executed by at least one processor to perform (see details). Figure 1 (Description) The function of electrostatic detection for cableway steel cables.

[0075] In this embodiment, the electrostatic detection system for the cableway steel cable can be divided into multiple functional modules according to its functions, such as... Figure 3As shown. The module referred to in this invention is a series of computer program segments that can be executed by at least one processor and perform a fixed function, and is stored in memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0076] The benchmark construction module is used to collect multimodal data of the cableway steel cable through a distributed sensor network under conditions without electrostatic risks; based on the multimodal data, a dynamic benchmark model that can characterize the mapping relationship between the real-time operating state and the normal charge range is trained. The real-time detection module is used to synchronously collect real-time multimodal data during cableway operation, and obtain the corresponding target normal charge range from the dynamic reference model based on the real-time multimodal data. The signal denoising module is used to perform adaptive denoising processing on the real-time mixed electrostatic signals in the real-time multimodal data in order to extract the denoised electrostatic signals. The risk determination module is used to compare the electrostatic charge corresponding to the electrostatic signal with the target normal charge range, and determine the risk level based on the comparison result.

[0077] Figure 4 The electrostatic detection method for cableway steel cables provided in this application embodiment can be applied to equipment. Those skilled in the art will understand that the equipment structure involved in the embodiments of this invention does not constitute a limitation on the equipment. The equipment may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, the equipment includes, but is not limited to, laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The equipment may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0078] The device 400 may include a processor 410, a memory 420, and a communication unit 430. These components communicate via one or more buses. Those skilled in the art will understand that the server structure shown in the figure does not constitute a limitation of the present invention. It may be a bus topology or a star topology, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0079] The memory 420 can be used to store execution instructions of the processor 410. The memory 420 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. When the execution instructions in the memory 420 are executed by the processor 410, the device 400 is able to perform some or all of the steps in the above method embodiments.

[0080] The processor 410 serves as the control center of the storage device, connecting various parts of the electronic device via various interfaces and lines. It executes software programs and / or modules stored in the memory 420, and calls data stored in the memory to perform various functions of the electronic device and / or process data. The processor can be composed of integrated circuits (ICs), such as a single packaged IC or multiple packaged ICs with the same or different functions connected together. For example, the processor 410 may consist only of a central processing unit (CPU). In this embodiment of the invention, the CPU may have a single processing core or include multiple processing cores.

[0081] The communication unit 430 is used to establish a communication channel, enabling the storage device to communicate with other devices. It can receive user data sent by other devices or send user data to other devices.

[0082] The present invention also provides a computer storage medium, wherein the computer storage medium may store a program, which, when executed, may include some or all of the steps provided in the embodiments of the present invention. The storage medium may be a magnetic disk, an optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0083] Those skilled in the art will clearly understand that the techniques in the embodiments of the present invention can be implemented using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, or any other medium capable of storing program code. It includes several instructions to cause a computer device (which may be a personal computer, a server, or a second device, network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0084] The same or similar parts between the various embodiments in this specification can be referred to mutually. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple, and the relevant parts can be referred to the description in the method embodiments.

[0085] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or modules may be electrical, mechanical, or other forms.

[0086] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0087] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0088] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the present invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should also be covered within the protection scope of the present invention.

Claims

1. A method of electrostatic detection of a cable of a ropeway, characterized in that, The method comprises the following steps: Under the condition of no static risk, multi-modal data of the cable of the cableway is collected by a distributed sensor network; Based on the multi-modal data, a dynamic reference model capable of representing the mapping relationship between the real-time running state and the normal charge range is trained; During the operation of the cableway, real-time multi-modal data is synchronously collected, and the corresponding target normal charge range is obtained from the dynamic reference model according to the real-time multi-modal data; The real-time mixed static signal in the real-time multi-modal data is adaptively denoised to extract the denoised static signal; The static charge corresponding to the static signal is compared with the target normal charge range, and the risk level is determined based on the comparison result.

2. The method of claim 1, wherein, Under the condition of no static risk, multi-modal data of the cable of the cableway is collected by a distributed sensor network, comprising: At least one of the following conditions is selected: high environmental humidity weather, rainy day or low-speed idle running state of the cableway, and data of the static signal, vibration signal, electromagnetic interference signal, environmental temperature and humidity, wind speed and cable running speed is continuously collected for 24 to 72 hours.

3. The method of claim 1, wherein, Based on the multi-modal data, a dynamic reference model capable of representing the mapping relationship between the working condition parameters and the normal charge range is trained, comprising: A sensor space-time graph is constructed, wherein the nodes are each sensor measuring point in the distributed sensor network, the node features are multi-modal data based on a time window, and the edges are defined according to the physical location relationship or signal correlation between the sensors; The space-time graph is input into a space-time graph attention network, and the low-dimensional space-time embedding features of each node are learned and output through the graph attention mechanism in the spatial dimension and the sequence model in the time dimension; Based on the low-dimensional space-time embedding features of all nodes, a deep embedding clustering algorithm is used for unsupervised clustering, wherein the clustering loss and the network reconstruction loss are jointly optimized; According to the clustering result, the corresponding normal charge range of each clustering cluster is calculated to generate the dynamic reference model.

4. The method of claim 3, wherein, According to the real-time multi-modal data, the corresponding target normal charge range is obtained from the dynamic reference model, comprising: The real-time space-time graph constructed based on the real-time multi-modal data is input into the trained space-time graph attention network encoder to obtain real-time low-dimensional space-time embedding features; The similarity between the real-time embedding features and the cluster center in the dynamic reference model is calculated; The normal charge range corresponding to the clustering cluster with the highest similarity is taken as the target normal charge range corresponding to the current real-time working condition.

5. The method of claim 1, wherein, The real-time mixed static signal in the real-time multi-modal data is adaptively denoised to extract the denoised static signal, comprising: The vibration signal and the electromagnetic interference signal are extracted from the real-time multi-modal data; The vibration signal and the electromagnetic interference signal are input as reference noise, and the signal component related to the reference noise is canceled from the real-time mixed static signal in real time by a normalized least mean square adaptive filter to output the denoised static signal.

6. The method of claim 1, wherein, The static charge corresponding to the static signal is compared with the target normal charge range, and the risk level is determined based on the comparison result. calculating a real-time electrostatic charge based on the denoised electrostatic signal; comparing the real-time electrostatic charge with a current normal electrostatic charge range obtained from a dynamic reference model; determining a risk level according to the comparison result according to the following rules: if the real-time electrostatic charge is within the normal range, determining as normal state; if the real-time electrostatic charge exceeds the normal range but is lower than a first threshold, or the charge accumulation rate exceeds a rate threshold, determining as attention state; if the real-time electrostatic charge reaches or exceeds the first threshold, or intermittent discharge pulses are monitored, determining as pre-warning state and triggering sound-light pre-warning; if the real-time electrostatic charge reaches or exceeds a second threshold higher than the first threshold, or continuous strong discharge is monitored, determining as alarm state and triggering the highest level alarm.

7. The method of claim 6, wherein, calculating a real-time electrostatic charge based on the denoised electrostatic signal, comprising: calculating the effective value of the voltage amplitude of the electrostatic signal after adaptive denoising processing according to a preset time window, and taking the effective value as the real-time electrostatic potential; calculating the real-time electrostatic charge based on the real-time electrostatic potential and the dynamic cable-to-ground capacitance value matched with the current working condition pre-established through simulation or experiment through the relationship Q=C×V; wherein Q is the charge amount, C is the dynamic capacitance value, and V is the real-time electrostatic potential.

8. An electrostatic detection system for a cable of a ropeway, characterized in that comprising: a reference modeling module for collecting multi-modal data of a cableway steel cable under a static electricity-free risk working condition through a distributed sensor network; training a dynamic reference model capable of representing the mapping relationship between the real-time running state and the normal charge amount range based on the multi-modal data; a real-time detection module for synchronously collecting real-time multi-modal data during cableway operation, and obtaining a corresponding target normal charge amount range from the dynamic reference model according to the real-time multi-modal data; a signal denoising module for adaptively denoising the real-time mixed electrostatic signal in the real-time multi-modal data to extract a denoised electrostatic signal; a risk determination module for comparing the electrostatic charge corresponding to the electrostatic signal with the target normal charge amount range, and determining a risk level based on the comparison result.

9. An electrostatic detection device for a cable of a ropeway, characterized in that comprising: a memory for storing a static electricity detection program of a cableway steel cable; a processor for implementing the steps of the static electricity detection method of the cableway steel cable according to any one of claims 1-7 when executing the static electricity detection program of the cableway steel cable.

10. A computer readable storage medium storing a computer program, characterized in that, The readable storage medium has stored thereon a static electricity detection program of a cableway steel cable, and the static electricity detection program of the cableway steel cable implements the steps of the static electricity detection method of the cableway steel cable according to any one of claims 1-7 when executed by a processor.

Images