Elevator operation quality monitoring system and monitoring method thereof

Abstract

The invention discloses an elevator running quality monitoring system and method, and the method comprises the steps: 1, obtaining the acceleration, current, vibration and position multi-mode signals of an elevator, and carrying out the time synchronization and physical correlation analysis, so as to convert a plurality of signals into a joint feature space; 2, taking the joint feature space as input, and improving the resolution of the multi-modal signal by using quantum calculation to obtain an enhanced signal; step 3, reconstructing a multi-mode signal of the elevator based on the enhanced signal, and generating a virtual mode to reflect a running track of the elevator in a constant speed state; and 4, on the basis of the generated virtual mode, the running state of the elevator is monitored in real time. According to the method, the precision problem of feature extraction in the constant-speed state is solved, tiny signal changes in the constant-speed state can be effectively captured, the monitoring precision of the elevator running state, the fault detection capacity and the early warning timeliness are improved, and the performance of an elevator monitoring system is remarkably enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator monitoring, and in particular to a monitoring system and a monitoring method for elevator operation quality. Background Art

[0002] With the rapid development of intelligent and automated technologies, elevators, as daily transportation tools, have gradually become an indispensable part of modern buildings. In order to ensure the safety and comfort of elevator operation, real-time monitoring of elevator operation status has become particularly important. Traditional elevator monitoring methods usually rely on single sensor data, such as acceleration, vibration, temperature, etc., to determine the operation status. However, during elevator operation, especially at a constant speed, various sensor data present a relatively stable state, which makes it difficult to capture small abnormal fluctuations or potential faults through feature extraction of a single signal, thus affecting the accurate judgment of elevator operation quality. Chinese patent CN114132810A proposes a solution that is applicable to all elevators and can quickly and accurately determine the operating status of the elevator. Although this elevator operating status monitoring method provides a more accurate way to monitor the status of the elevator, there are still some technical challenges and potential problems. The acceleration under the uniform speed state should theoretically be zero, but when the elevator is running, especially under the uniform speed state, although the acceleration and deceleration actions of the elevator itself have ended, due to mechanical wear, uneven load or minor internal faults of the system, the running quality of the elevator may change slightly. These minor fluctuations usually affect signals such as acceleration, vibration, and current, and these changes are not easy to detect under the uniform speed state. Traditional signal processing and feature extraction methods often cannot fully identify these details. Therefore, the accuracy of feature extraction under the uniform speed state is one of the key difficulties in the current elevator operation quality monitoring method. Summary of the invention

[0003] In order to solve the problem of accuracy of feature extraction under a uniform speed state, the present invention provides a monitoring system and a monitoring method for elevator operation quality.

[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] On the one hand, the present invention discloses a monitoring system and a monitoring method for elevator operation quality, comprising:

[0006] Step 1: Obtain elevator acceleration, current, vibration and position multimodal signals, and perform time synchronization and physical correlation analysis to convert multiple signals into a joint feature space;

[0007] Step 2: Using the joint feature space as input, quantum computing is used to improve the resolution of the multimodal signal to obtain an enhanced signal;

[0008] Step 3: Reconstruct the multimodal signal of the elevator based on the enhanced signal and generate a virtual model to reflect the running trajectory of the elevator under a uniform speed state;

[0009] Step 4: Based on the generated virtual model, monitor the operating status of the elevator in real time.

[0010] Further: Step 1 comprises:

[0011] First, the acceleration, current, vibration and position data are standardized to eliminate the dimension differences and scale deviations of different sensor signals;

[0012] Time synchronization and alignment of the standardized data;

[0013] Map the subsequent signal to an embedding space E i ;

[0014] For each sensor signal E i , by calculating the weighting coefficient α ij , get the dependency between each pair of signals, and perform weighted summation on the signals of different sensors to get the final joint feature representation H t :

[0015] The obtained joint feature representation H t , through the multi-layer perceptron nonlinear transformation, the final feature representation Z is obtained t .

[0016] Further: Step 2 includes:

[0017] The feature is represented by Z t As input, it is mapped to the frequency domain through quantum Fourier transform

[0018] Apply quantum noise filter to remove noise and enhance the accuracy of signal to obtain filtered signal Z filtered ;

[0019] By taking the filtered signal Z filtered Mapping to quantum space and using the high-dimensional characteristics of quantum states to enhance the resolution of the signal, forming a strong signal Z enhance .

[0020] Further: Step 3 comprises:

[0021] The enhanced signal Z enhance Mapping into holographic space;

[0022] After being mapped to the holographic space, it is restored using a holographic reconstruction algorithm;

[0023] By analyzing the restored signal, a virtual mode Z of the elevator running at a constant speed is established. virtual .

[0024] Further: Step 4 includes:

[0025] The feature is represented by Z t With virtual mode Z virtual The deviation is measured by weighted Euclidean distance:

[0026]

[0027] Among them, Z t,i and Z virtual,i are the values ​​of the real-time signal and the virtual mode on the i-th sensor, respectively, and w i is the weight coefficient of acceleration, current, vibration and position, indicating the importance of each signal in the system;

[0028] When the deviation δ t When the preset threshold ∈ is exceeded, the elevator operation status is judged to be abnormal, and an early warning is triggered.

[0029] On the other hand, the present invention discloses a method for monitoring the operation quality of an elevator, comprising:

[0030] Multimodal signal fusion and feature space construction module: obtains elevator acceleration, current, vibration and position multimodal signals, and performs time synchronization and physical correlation analysis to convert multiple signals into a joint feature space;

[0031] Quantum-enhanced signal resolution and abnormal pattern recognition module: using the joint feature space as input, quantum computing is used to improve the resolution of multimodal signals to obtain enhanced signals;

[0032] Holographic acceleration data reconstruction and virtual mode generation module: reconstructs the multimodal signal of the elevator based on the enhanced signal and generates a virtual mode to reflect the running trajectory of the elevator under uniform speed state;

[0033] Real-time warning module: Based on the generated virtual model, it monitors the operating status of the elevator in real time.

[0034] Compared with the prior art, the present invention has the following technical advances:

[0035] The present invention can more comprehensively capture the operating characteristics of the elevator under uniform speed state by fusing the four sensor signals of acceleration, current, vibration and position of the elevator into a unified feature space. Compared with a single signal, multimodal data fusion enables small changes of different signals to complement each other, thereby greatly improving the accuracy of signal feature extraction under uniform speed state. In particular, vibration signals and current signals have smaller fluctuations under uniform speed state, but they can reveal potential problems in the system, such as mechanical wear or uneven load. The introduction of quantum computing can accurately capture small signal changes under uniform speed state, making up for the traditional signal processing method's neglect of small fluctuations under uniform speed state. Therefore, the quantum enhanced signal not only maintains high accuracy, but also can reveal abnormal changes that are difficult to detect by traditional methods, further improving the quality of feature extraction under uniform speed state.

[0036] The present invention generates a more realistic virtual model with the help of holographic acceleration data reconstruction technology, so that the running trajectory of the elevator under the uniform speed state can be accurately reproduced, and the slight changes in the signal under the uniform speed state can be accurately identified by comparing with the virtual model, effectively avoiding the situation where the conventional feature extraction method cannot fully identify the slight fluctuations under the uniform speed state. The present invention can timely identify abnormal fluctuations under the uniform speed state through deviation analysis between the virtual model and the real-time signal. Due to the fusion of multimodal data and high-precision reconstruction of the signal, the system can detect potential faults earlier and more accurately, thereby early warning of abnormal conditions in the elevator operation. Traditional methods often rely on local features or fluctuations of a single signal, while this method can comprehensively and accurately identify slight abnormal changes under the uniform speed state by combining feature extraction of multiple signals, ensuring the long-term stability of the elevator operation quality.

[0037] In summary, the present invention overcomes the accuracy problem of feature extraction under uniform speed state, can effectively capture tiny signal changes under uniform speed state, improves the monitoring accuracy of elevator operation status, fault detection capability and timeliness of early warning, and significantly enhances the performance of elevator monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0039] In the attached picture:

[0040] Figure 1 This is a flow chart of Embodiment 1 of the present invention;

[0041] Figure 2 This is a system structure diagram of Example 2 of the present invention. DETAILED DESCRIPTION

[0042] The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0043] Example 1

[0044] like Figure 1 As shown, a method for monitoring the operation quality of an elevator comprises:

[0045] Step 1: Obtain elevator acceleration, current, vibration and position multimodal signals, and perform time synchronization and physical correlation analysis to convert multiple signals into a joint feature space;

[0046] Step 2: Using the joint feature space as input, quantum computing is used to improve the resolution of the multimodal signal to obtain an enhanced signal;

[0047] Step 3: Reconstruct the multimodal signal of the elevator based on the enhanced signal and generate a virtual model to reflect the running trajectory of the elevator under a uniform speed state;

[0048] Step 4: Based on the generated virtual model, monitor the operating status of the elevator in real time.

[0049] Specifically, step 1 includes:

[0050] 1. Signal acquisition and fusion:

[0051] First, multimodal signals of acceleration, current, vibration and position are obtained, including using an acceleration sensor to obtain the acceleration data of the elevator car; using a current sensor to monitor the current fluctuations of the elevator motor, and obtaining information such as the motor operating load and efficiency from the current signal; using a vibration sensor to obtain the mechanical vibration signal generated by the elevator equipment (such as the motor, guide rails, base, etc.) during operation, which is used to identify the mechanical state inside the elevator; using a position sensor to monitor the current position and motion state of the elevator car, especially to accurately monitor the displacement changes of the car under a uniform speed state.

[0052] 2. Data preprocessing and standardization

[0053] In order to effectively fuse data from different sensors, the acceleration, current, vibration and position data must first be standardized to eliminate the dimensional differences and scale deviations of different sensor signals. The signal standardization formula is: Among them, X is the original signal data, μ X is the mean of the data, σ X is the standard deviation of the data. Through standardization, the data of each sensor can be mapped to a unified scale, so that they can be effectively jointly modeled in subsequent steps.

[0054] 3. Time synchronization and data alignment

[0055] Since the various sensors of the elevator may collect data at different time steps, these data need to be synchronized and aligned. In order to ensure that the data of each time step can be aligned and processed, the time window alignment technology is used to crop the data of all sensors at the same time interval and set a time window T. w and sampling interval Δt, all sensor signals at each time point can be expressed as X t ,Through such a time window, it is ensured that all sensor data at each time step t can be collected at the same time point and used for the subsequent model training and prediction.

[0056] 4. Multimodal feature fusion and joint modeling

[0057] The fusion of multimodal data is achieved in this embodiment through a multimodal deep learning model, which performs weighted aggregation on signals from different sensors through a self-attention mechanism to form a joint feature representation.

[0058] 4.1 Feature Mapping and Multimodal Input Encoding

[0059] Map the normalized signal of each sensor to an embedding space E i (i.e., vector representation of different sensor data), the mapping formula is: E i =f(X i ) where f represents a fully connected network used to transform each sensor signal X i Mapped to a high-dimensional feature space.

[0060] 4.2 Self-Attention Mechanism and Weighted Aggregation

[0061] Next, the self-attention mechanism is used to weight the data from different sensors. For each sensor signal E i , by calculating the weighting coefficient α ij (i.e., the correlation weights between the signals), we can get the dependency between each pair of signals:

[0062]

[0063] Among them, E i and E j are the embedding vectors from sensor i and sensor j respectively, which are high-dimensional feature representations obtained by normalizing each sensor signal and passing it through a mapping function (such as a fully connected layer or other neural network layer).

[0064] and It is the transpose of the embedded vectors, that is, these vectors are regarded as column vectors and transformed into row vectors through the transpose operation. The purpose of the transpose operation is to facilitate the calculation of the inner product (dot product) to measure the correlation or similarity of the two sensor signals.

[0065] Inner Product The similarity between sensor i and sensor j is calculated. If the embedding vectors of the two are similar, the inner product value is large; if their signal characteristics are very different, the inner product value is small.

[0066] The inner product value is subjected to exponential mapping in order to amplify the influence of the signal with greater similarity on the subsequent weighting and to compress the influence of the signal with less similarity.

[0067] Finally, the normalization term Make sure that the sum of all similarities is 1 so that the influence of each sensor signal is not too large or too small when combined.

[0068] This fusion method based on the self-attention mechanism can automatically learn the relationship and dependency between the data of each sensor, and then generate a more accurate joint feature space, providing a solid data foundation for the subsequent analysis of the elevator operation status. Next, the signals of different sensors are weighted and summed to obtain the final joint feature representation:

[0069]

[0070] Through this weighted aggregation, the system can automatically learn the relationships between different sensor data, and thus form a joint feature space containing multimodal information.

[0071] 4.3 Multi-layer Perceptron Feature Transformation

[0072] The obtained joint feature H t , further nonlinear transformation is performed through multi-layer perceptron to enhance the expressive power of features. Assuming that after L layers of MLP transformation, the final feature representation Z is obtained t ,Through the transformation of the multi-layer perceptron, the system can extract more ,discriminative features suitable for subsequent classification or ,regression tasks.

[0073] 5. Construction and output of joint feature space

[0074] Through the above multimodal fusion and weighted aggregation, we finally get a joint feature space Z that integrates all sensor data. t ,This space can fully express various information about the elevator’s operating status and provide accurate feature input for subsequent analysis, fault detection, and anomaly identification.

[0075] Finally, by combining the feature space Z t Output the elevator's operating status or other monitoring indicators. This output not only includes the elevator's precise acceleration data at a constant speed, but also integrates other signal information such as current, vibration, and position, greatly improving the system's robustness and accuracy in complex situations.

[0076] Specifically, step 2 includes:

[0077] The joint feature space Z obtained by multimodal data fusion t Based on the above, quantum computing technology is used to enhance the resolution of the signal, especially the ability to detect small changes in a uniform speed state. Then, the quantum support vector machine (QSVM) is used to identify abnormal patterns in the processed multimodal data, so as to detect potential faults in the elevator operation process in real time. This step ensures that the elevator operation quality can be accurately identified and judged in a uniform speed state.

[0078] Quantum computing can significantly improve the efficiency and accuracy of signal analysis through features such as parallel processing and quantum state superposition. In this step, quantum computing is mainly used to optimize the data of acceleration, current, vibration, and position sensors so that small changes can be more keenly identified at a constant speed.

[0079] 1 Quantum Signal Processing

[0080] First, the joint feature space Z obtained by multimodal fusion in the previous step is t As input, quantum computing is used to process the signal at each moment. In order to improve the resolution of the signal, quantum Fourier transform is used to perform frequency domain analysis on the signal of each sensor to obtain a higher precision spectrum.

[0081] Assume Z t is a joint feature at a certain point in time, which can be mapped to the frequency domain through quantum Fourier transform

[0082] 2 Quantum Noise Filtering

[0083] Elevator signals often contain noise, especially at a constant speed, where small changes in the signal are more susceptible to noise. In order to improve signal quality, quantum noise filters are used to remove noise and enhance signal accuracy. Quantum noise filters effectively suppress unnecessary noise and retain key components in the signal through quantum superposition and quantum interference. filtered is the signal after filtering.

[0084] 3. Quantum Resolution Enhancement

[0085] During the processing, quantum computing uses its ultra-high resolution capability to improve the resolution of the signal through the superposition effect of quantum bits to accurately capture the tiny acceleration changes of the elevator at a constant speed. This process can enhance the resolution of the signal by mapping the signal to quantum space and utilizing the high-dimensional characteristics of the quantum state to form a high-precision signal representation Z. enhance , Z enhance is the enhanced signal.

[0086] Specifically, step 3 includes:

[0087] The goal of step 3 is to obtain the high-precision multimodal signal Z based on step 2. enhance The holographic acceleration data reconstruction technology is used to reconstruct the elevator's multimodal signals (including acceleration, current, vibration and position) and generate a virtual model. This virtual model can truly reflect the running trajectory of the elevator under a uniform speed state, and then match it with historical data to identify abnormal fluctuations that may occur under a uniform speed state. Through this process, the extraction accuracy of acceleration features can be effectively improved, thereby optimizing the judgment of the elevator's operating status.

[0088] 1. Reconstruction of holographic acceleration data

[0089] The core idea of ​​holographic acceleration data reconstruction is to use the existing high-precision multimodal signals to restore a more comprehensive and accurate elevator operation status in the holographic space. By building a holographic data reconstruction model, the model can combine the elevator's multimodal sensor signals to reconstruct complete information such as acceleration, current, vibration and position data.

[0090] 1.1 Signal mapping to holographic space

[0091] First, the enhanced signal Z obtained in step 2 is enhance Mapped to the holographic space, assuming that the joint feature of the original signal is Z enhance , through a holographic mapping function H(·), Z enhance Mapped to a high-dimensional holographic space H(Z enhance ).

[0092] 1.2 Holographic signal reconstruction

[0093] Once the signal is mapped to the holographic space, it can be restored using a holographic reconstruction algorithm. The goal of holographic reconstruction is to reconstruct the original signal of the elevator at a constant speed based on the information in the holographic space.

[0094] 2. Virtual mode generation

[0095] The generation of the virtual model is completed by simulating the running trajectory of the elevator at a constant speed. In steps 1 and 2, high-precision signals from each sensor have been obtained, so the generation process of the virtual model is based on these signal data and accurately simulates the running trajectory of the elevator at a constant speed by establishing a mathematical model.

[0096] 2.1 Elevator operation trajectory modeling

[0097] By reconstructing the signal Z enhance After analysis, we can establish a mathematical model of the elevator's uniform speed operation state. Considering the data change trend of each sensor under the uniform speed state, assuming that the motion trajectory of the elevator under the uniform speed state can be described by the association between position and acceleration data, we can construct a trajectory generation formula based on the time series model:

[0098] Z virtual =M(Z enhance,t )

[0099] Among them, Z virtual is the generated virtual mode, M(·) is the time series modeling function describing the elevator running trajectory under uniform speed state, and t is the time series.

[0100] Specifically, step 4 includes:

[0101] The goal of step 4 is to pass the virtual pattern Z generated in step 3 virtual , based on real-time monitoring of the elevator's operating status, combined with deep reinforcement learning (DRL) for intelligent decision-making. When the elevator's multimodal signals (including acceleration, current, vibration, and position) deviate from the normal virtual mode, the system can automatically trigger a fault warning and provide maintenance suggestions. In addition, the system uses an adaptive optimization algorithm to adjust the monitoring thresholds and parameters based on real-time data to ensure high-precision monitoring effects under different operating conditions.

[0102] 1. Real-time signal monitoring and virtual mode deviation calculation

[0103] First, it is necessary to monitor the multimodal signals of the elevator in real time, including acceleration, current, vibration, and position. These signals are highly synchronized in time and must be synchronized with the virtual mode Z generated in step 3. virtual Compare and detect whether there is deviation. Assume that the multimodal signal received by the system in real time is Z t , then the deviation can be measured by weighted Euclidean distance, the formula is as follows:

[0104]

[0105] Among them, Z t,i and Z virtual,iare the values ​​of the real-time signal and the virtual mode on the i-th sensor, respectively, and w i is the weight coefficient of acceleration, current, vibration and position, indicating the importance of each signal in the system

[0106] Among them, Z t,i and Z virtual,i are the values ​​of the real-time signal and the virtual mode on the i-th sensor, respectively, and w i is the weight coefficient of acceleration, current, vibration and position, indicating the importance of each signal in the system.

[0107] When the deviation δ t When the preset threshold ∈ is exceeded, the elevator operation status is judged to be abnormal, and an early warning is triggered.

[0108] 2. Deep reinforcement learning (DRL) intelligent decision-making mechanism

[0109] When an anomaly that deviates from the virtual mode is detected, appropriate countermeasures must be taken. At this time, deep reinforcement learning (DRL) is used to achieve intelligent decision-making. Through the DRL model, decision-making strategies can be gradually learned and optimized through multiple experiments and feedback. The core idea of ​​DRL is that the intelligent agent continuously optimizes the strategy through interaction with the environment (i.e. feedback from real-time monitoring data), thereby providing the best fault diagnosis and maintenance suggestions.

[0110] 2.1 Reinforcement Learning Model Definition

[0111] In this process, it can be regarded as a reinforcement learning agent, and the environment consists of the real-time operation data and status of the elevator. Assume that the operation status of the elevator is composed of s t Indicates that the action space is A, including triggering fault warning, performing maintenance, adjusting monitoring thresholds, etc. Through interaction with the environment, the agent can determine the current state s t Select an action t , and according to the reward signal r fed back by the system t Update the policy.

[0112] The goal of reinforcement learning is to maximize long-term rewards. The agent's strategy is learned by the deep Q network (DQN). Taking DQN as an example, the formula for updating the Q value is as follows:

[0113]

[0114] Among them, α is the learning rate, γ is the discount factor, and r t is the immediate reward, Q(s t ,a t ) is the current state s t and action a t Q value.

[0115] 2.2 Action Decision and Maintenance Suggestions

[0116] According to the current status t and action a t , the agent chooses the best action based on the learned strategy. For example, when the deviation δ t When it is large, the system may select the "fault warning" action and give maintenance suggestions; if the selected action is "fault warning", the warning will be triggered and specific maintenance suggestions will be output, which may include inspection of elevator mechanical parts, current fluctuation analysis, vibration fault troubleshooting, etc.

[0117] Example 3

[0118] This embodiment discloses a method for monitoring elevator operation quality, including:

[0119] Multimodal signal fusion and feature space construction module: obtains elevator acceleration, current, vibration and position multimodal signals, and performs time synchronization and physical correlation analysis to convert multiple signals into a joint feature space;

[0120] Quantum-enhanced signal resolution and abnormal pattern recognition module: using the joint feature space as input, quantum computing is used to improve the resolution of multimodal signals to obtain enhanced signals;

[0121] Holographic acceleration data reconstruction and virtual mode generation module: reconstructs the multimodal signal of the elevator based on the enhanced signal and generates a virtual mode to reflect the running trajectory of the elevator under uniform speed state;

[0122] Real-time warning module: Based on the generated virtual model, it monitors the operating status of the elevator in real time.

[0123] The module in the embodiment 2 is used to implement the function in the embodiment 1. This embodiment can be implemented by a system, the system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the prediction model construction method for postoperative cognitive decline in elderly patients according to the embodiment 1 of the present application is implemented. The system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, whose settings and functions are known in the art, so they are not repeated here.

[0124] In the present application, the aforementioned memory may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, system, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in this application may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0125] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the claims of the present invention.

Claims

1. A method for monitoring elevator operation quality, characterized in that: include: Step 1: Obtain elevator acceleration, current, vibration and position multimodal signals, and perform time synchronization and physical correlation analysis to convert multiple signals into a joint feature space; Step 2: Using the joint feature space as input, quantum computing is used to improve the resolution of the multimodal signal to obtain an enhanced signal; Step 3: Reconstruct the multimodal signal of the elevator based on the enhanced signal and generate a virtual model to reflect the running trajectory of the elevator under a uniform speed state; Step 4: Based on the generated virtual model, monitor the operating status of the elevator in real time.

2. A method for monitoring elevator operation quality according to claim 1, characterized in that: The step 1 comprises: First, the acceleration, current, vibration and position data are standardized to eliminate the dimension differences and scale deviations of different sensor signals; Time synchronization and alignment of the standardized data; Map the subsequent signal to an embedding space E i ; For each sensor signal E i , by calculating the weighting coefficient α ij , get the dependency between each pair of signals, and perform weighted summation on the signals of different sensors to get the final joint feature representation H t : The obtained joint feature representation H t , through the multi-layer perceptron nonlinear transformation, the final feature representation Z is obtained t .

3. The method for monitoring elevator operation quality according to claim 2, characterized in that: The step 2 comprises: The feature is represented by Z t As input, it is mapped to the frequency domain through quantum Fourier transform Apply quantum noise filter to remove noise and enhance the accuracy of signal to obtain filtered signal Z filtered ; By taking the filtered signal Z filtered Mapping to quantum space and using the high-dimensional characteristics of quantum states to enhance the resolution of the signal, forming a strong signal Z enhance .

4. The method for monitoring elevator operation quality according to claim 3, characterized in that: The step 3 comprises: The enhanced signal Z enhance Mapping into holographic space; After being mapped to the holographic space, it is restored using a holographic reconstruction algorithm; By analyzing the restored signal, a virtual mode Z of the elevator running at a constant speed is established. virtual .

5. The method for monitoring elevator operation quality according to claim 4, characterized in that: The step 4 comprises: The feature is represented by Z t With virtual mode Z virtual The deviation is measured by weighted Euclidean distance: Among them, Z t,i and Z virtual,i are the values ​​of the real-time signal and the virtual mode on the i-th sensor, respectively, and w i is the weight coefficient of acceleration, current, vibration and position, indicating the importance of each signal in the system; When the deviation δ t When the preset threshold ∈ is exceeded, the elevator operation status is judged to be abnormal, and an early warning is triggered.

6. A method for monitoring the running quality of an elevator, characterized in that: include: Multimodal signal fusion and feature space construction module: obtains elevator acceleration, current, vibration and position multimodal signals, and performs time synchronization and physical correlation analysis to convert multiple signals into a joint feature space; Quantum-enhanced signal resolution and abnormal pattern recognition module: using the joint feature space as input, quantum computing is used to improve the resolution of multimodal signals to obtain enhanced signals; Holographic acceleration data reconstruction and virtual mode generation module: reconstructs the multimodal signal of the elevator based on the enhanced signal and generates a virtual mode to reflect the running trajectory of the elevator under uniform speed state; Real-time warning module: Based on the generated virtual model, it monitors the operating status of the elevator in real time.

Images