Method and System for Monitoring the Operating Status of Park Facilities Based on Multimodal Sensing

Abstract

The present invention relates to the technical field of multi-modal perception monitoring, specifically including a method and system for monitoring the operation status of park facilities based on multi-modal perception, which includes: collecting multi-modal perception data, generating a fusion matrix, extracting operation features, constructing a facility status evolution map, configuring a monitoring period, identifying potential failure modes, matching with historical data to optimize monitoring rules, and outputting a monitoring report. It solves the technical problem that the potential failure features of cross-modal associations cannot be effectively extracted, affecting the accuracy and comprehensiveness of failure identification, and realizes the construction of a facility status evolution map. By combining the time prediction of the health index and the spatial annotation of the failure risk probability, latent failures are identified in advance. At the same time, through multi-modal data retrospective analysis and state transition feedback optimization, the sensitivity of failure feature extraction is improved, and incremental learning and model iteration are used to continuously optimize the adaptive association model, achieving the technical effect of improving the accuracy and comprehensiveness of failure identification.

Description

Technical Field

[0001] The present invention relates to the technical field of multimodal perception monitoring, and specifically relates to a method and system for monitoring the operating status of park facilities based on multimodal perception. Background Technique

[0002] With the continuous development of technologies such as the Internet of Things, big data, and artificial intelligence, the technology for monitoring the operating status of park facilities is also constantly advancing. Through the integration of advanced information technologies such as GIS (Geographic Information System), Internet of Things technology, BIM technology, 5G technology, cloud computing technology, big data technology, and artificial intelligence technology, smart parks have realized the intelligent transformation and upgrading of infrastructure, public safety, energy management, environmental protection, etc. within the park, providing more comprehensive and accurate data support for monitoring the operating status of park facilities.

[0003] However, existing fault recognition rules are mostly based on static thresholds or single indicators (such as health index), and cannot adaptively adjust the monitoring period and threshold range according to dynamic parameters such as environmental temperature and load rate, resulting in false alarms or missed alarms. Therefore, how to effectively apply multimodal perception technology to the monitoring of the operating status of park facilities and improve the accuracy of monitoring has become an urgent problem to be solved.

[0004] In summary, there is a technical problem in the prior art that the pattern matching mechanism of the historical fault database is single, and it is unable to effectively extract potential fault features with cross-modal associations, affecting the accuracy and comprehensiveness of fault recognition. Summary of the Invention

[0005] This application provides a system for monitoring the operating status of park facilities based on multimodal perception, aiming to solve the technical problem in the prior art that the pattern matching mechanism of the historical fault database is single, and it is unable to effectively extract potential fault features with cross-modal associations, affecting the accuracy and comprehensiveness of fault recognition.

[0006] In view of the above problems, the technical solution of this application is as follows:

[0007] On the one hand, the present application provides a method for monitoring the operating status of park facilities based on multimodal perception. The method includes: collecting multimodal perception data of each park facility in the target park, generating a fused data matrix through alignment processing, and inputting it into a pre-trained adaptive correlation model to extract facility operation features; constructing a facility status evolution map based on the facility operation features, including a facility health index, an energy consumption anomaly coefficient, and a failure risk probability; configuring a first monitoring period according to the facility health index and the energy consumption anomaly coefficient in the facility status evolution map, and identifying a first type of potential failure mode under the initial failure identification rule; configuring a second monitoring period according to the failure risk probability and the energy consumption anomaly coefficient in the facility status evolution map, and identifying a second type of potential failure mode under the initial failure identification rule; performing pattern matching on the first monitoring period and the first type of potential failure mode, and the second monitoring period and the second type of potential failure mode with a historical failure database, optimizing the monitoring period and the failure identification rule of each park facility according to the matching result, and outputting a failure monitoring report.

[0008] On the other hand, the present application provides a system for monitoring the operating status of park facilities based on multimodal perception. The system includes: a data collection module for collecting multimodal perception data of each park facility in the target park, generating a fused data matrix through alignment processing, and inputting it into a pre-trained adaptive correlation model to extract facility operation features; a map construction module for constructing a facility status evolution map based on the facility operation features, including a facility health index, an energy consumption anomaly coefficient, and a failure risk probability; a first identification module for configuring a first monitoring period according to the facility health index and the energy consumption anomaly coefficient in the facility status evolution map, and identifying a first type of potential failure mode under the initial failure identification rule; a second identification module for configuring a second monitoring period according to the failure risk probability and the energy consumption anomaly coefficient in the facility status evolution map, and identifying a second type of potential failure mode under the initial failure identification rule; a pattern matching module for performing pattern matching on the first monitoring period and the first type of potential failure mode, and the second monitoring period and the second type of potential failure mode with a historical failure database, optimizing the monitoring period and the failure identification rule of each park facility according to the matching result, and outputting a failure monitoring report.

[0009] In summary, one or more technical solutions provided in the present application achieve the construction of a facility status evolution map, combine the time prediction of the health index and the spatial annotation of the failure risk probability to identify latent failures in advance. At the same time, through multimodal data backtracking analysis and state transition feedback optimization, the sensitivity of failure feature extraction is improved. By using incremental learning and model iteration, the adaptive correlation model is continuously optimized, and the technical effects of improving the accuracy and comprehensiveness of failure identification are achieved. Description of the Drawings

[0010] Figure 1 This application provides a flow schematic diagram of a method for monitoring the operation status of park facilities based on multi-modal perception;

[0011] Figure 2 This application provides a structural schematic diagram of a system for monitoring the operation status of park facilities based on multi-modal perception.

[0012] Explanation of reference numerals: data acquisition module M100, atlas construction module M200, first recognition module M300, second recognition module M400, pattern matching module M500. Detailed implementation manners

[0013] Embodiment 1

[0014] The following specifically describes this application with reference to the accompanying drawings. As Figure 1 shown, this application provides a method for monitoring the operation status of park facilities based on multi-modal perception. Among them, the method includes:

[0015] S1: Collect multi-modal perception data of each park facility in the target park, generate a fusion data matrix through alignment processing, and input it into a pre-trained adaptive association model to extract facility operation features; S2: Based on the facility operation features, construct a facility status evolution atlas, including a facility health index, an energy consumption anomaly coefficient, and a failure risk probability.

[0016] Specifically, multi-modal perception data refers to data related to the operation status of park facilities collected by a variety of different types of sensors, including operation voiceprint signals (sound features collected by acoustic sensors), infrared thermal imaging atlases (temperature distribution images collected by infrared thermal imaging devices), and current harmonic features (harmonic components in the current collected by current sensors); since the data acquisition time, frequency, and method of different modalities may vary, synchronization processing is performed through time axis alignment operations to ensure consistency in time. For example, the sampling time windows of the operation voiceprint signal, infrared thermal imaging atlas, and current harmonic features are aligned to eliminate phase delay errors; the multi-modal data after alignment processing is integrated into a unified data structure to form a fusion data matrix, which contains feature information from different modalities.

[0017] The pre-trained adaptive association model is obtained by pre-training with a large amount of historical data. The pre-trained adaptive association model can dynamically learn the cross-modal correlation between multi-modal data, that is, the mutual relationship between different modal data. For example, it can learn the correlation between the anomalies in the running voiceprint signal and the current harmonic characteristics, so as to extract the facility operation characteristics more accurately. The facility operation characteristics refer to the key information that can reflect the facility operation state extracted from the fusion data matrix, including the vibration frequency of the equipment, the temperature change rate, the amplitude of the current harmonics, etc., which is the basis for constructing the facility state evolution map. The facility state evolution map includes the facility health index (a quantitative index reflecting the overall health status of the equipment), the energy consumption anomaly coefficient (an index measuring whether the equipment energy consumption is abnormal), and the failure risk probability (predicting the possibility of equipment failure). The facility state evolution map is presented in the way of time series analysis and spatial topology annotation, and can intuitively display the operation state of the facility and its change trend.

[0018] First, collect the multi-modal perception data of each facility in the target park, including the running voiceprint signal, the infrared thermal imaging map, and the current harmonic characteristics, perform synchronous processing, eliminate the phase delay error between different modal data, and map it into the feature space to generate a fusion data matrix. Next, input the fusion data matrix into the pre-trained adaptive association model to extract the facility operation characteristics and construct the facility state evolution map, including the facility health index, the energy consumption anomaly coefficient, and the failure risk probability.

[0019] By collecting multi-modal perception data and performing alignment processing, it can comprehensively and accurately reflect the operation state of the facility. The pre-trained adaptive association model can dynamically learn the correlation between multi-modal data and extract more representative operation characteristics. The facility state evolution map provides a basis for subsequent monitoring cycle configuration and fault mode recognition. For example, the facility health index can be used to evaluate the overall health status of the equipment, the energy consumption anomaly coefficient can timely detect the energy consumption anomaly of the equipment, and the failure risk probability can early warn of potential failure risks, enabling the monitoring system to more comprehensively and accurately master the operation state of the facility and provide strong support for subsequent fault diagnosis and optimization.

[0020] S3: Configure the first monitoring cycle according to the facility health index and the energy consumption anomaly coefficient in the facility state evolution map, and identify a class of potential fault modes under the initial fault identification rule; S4: Configure the second monitoring cycle according to the failure risk probability and the energy consumption anomaly coefficient in the facility state evolution map, and identify a second class of potential fault modes under the initial fault identification rule; S5: Match the first monitoring cycle and the first class of potential fault modes, the second monitoring cycle and the second class of potential fault modes with the historical fault database, optimize the monitoring cycle and the fault identification rule of each park facility with the matching result, and output a fault monitoring report.

[0021] Specifically, the monitoring period refers to the time interval for monitoring the operating status of the park facilities. Depending on the health status and failure risk of the facilities, the monitoring period can be dynamically adjusted. For example, for park facilities with a higher failure risk, the monitoring period will be shortened to obtain data more frequently and detect potential problems in a timely manner; the initial failure identification rule is used to preliminarily determine whether there are potential failures in the facilities, usually formulated based on the normal operating characteristics and historical failure data of the facilities. For example, when the facility health index is lower than a certain threshold or the energy consumption anomaly coefficient exceeds a certain range, it may trigger failure identification.

[0022] A type of potential failure mode refers to the failure types identified through the first monitoring period and the initial failure identification rule, which are usually closely related to the facility health index and the energy consumption anomaly coefficient, including mechanical wear, circuit overload, etc., and are a direct reflection of the current operating status of the facilities; a type II potential failure mode refers to the failure types identified through the second monitoring period and the initial failure identification rule, which are usually related to the failure risk probability and the energy consumption anomaly coefficient, including abnormal heat dissipation, harmonic distortion, etc., and pay more attention to the potential risks of the facilities and possible problems in the future; pattern matching refers to comparing and analyzing the currently identified type of potential failure mode and type II potential failure mode with the known failure modes in the historical failure database respectively to determine the similarity between the type of potential failure mode and type II potential failure mode; through pattern matching, the accuracy of the current failure mode can be verified, and a basis can be provided for the optimization of subsequent monitoring periods and failure identification rules; the failure monitoring report summarizes the operating status of the facilities, the identified potential failure modes, the optimized monitoring periods and failure identification rules, etc., and provides support for the maintenance and management of park facilities.

[0023] Dynamically configure the first monitoring period according to the facility health index and the energy consumption anomaly coefficient in the facility status evolution map. For example, when the facility health index is low or the energy consumption anomaly coefficient is high, shorten the monitoring period to monitor the facility status more frequently. Under the initial failure identification rule, identify a type of potential failure mode, such as mechanical wear or circuit overload, through real-time monitoring data; dynamically configure the second monitoring period according to the failure risk probability and the energy consumption anomaly coefficient in the facility status evolution map. The second monitoring period pays more attention to the potential risks of the facilities and identifies a type II potential failure mode, such as abnormal heat dissipation or harmonic distortion, through the initial failure identification rule.

[0024] Match the first monitoring period and a type of potential failure mode, the second monitoring period and a second type of potential failure mode with the historical failure database. Through pattern matching, verify the accuracy of the current failure mode, and optimize the monitoring period and failure identification rules according to the matching results. For example, if the current failure mode is highly similar to a certain failure mode in the historical database, the monitoring period can be adjusted to more accurately track the development of this failure mode; generate a failure monitoring report, summarizing the operating status of the facilities, the identified potential failure modes, the optimized monitoring period and failure identification rules, etc., to provide decision-making support for the maintenance and management of the park facilities.

[0025] By dynamically configuring the monitoring period and identifying potential failure modes, the system can flexibly adjust the monitoring strategy according to the actual operating status of the facilities, thereby improving the accuracy and efficiency of failure identification. For example, the identification of a type of potential failure mode focuses on the direct reflection of the current operating status, while the identification of a second type of potential failure mode pays more attention to the risks that may occur in the future. Through pattern matching, the system can use historical data to optimize the monitoring period and failure identification rules, forming a closed loop of "perception - diagnosis - optimization", improving the intelligent level of the system. The output failure monitoring report provides comprehensive and accurate information support for the maintenance and management of the park facilities, ensuring the safe operation of the facilities.

[0026] Furthermore, the method of this application includes:

[0027] The multi-modal perception data includes operating acoustic fingerprint signals, infrared thermal imaging maps, and current harmonic characteristics; in the power supply circuits corresponding to each park facility in the target park, deploy non-intrusive current sensors to obtain current harmonic characteristics including odd harmonic amplitudes and phase offsets; obtain the surface temperature field of the park facilities to generate infrared thermal imaging maps with timestamps; use a distributed acoustic sensor array to collect operating acoustic fingerprint signals.

[0028] Specifically, multi-modal perception data refers to data collected through a variety of different types of sensors, used to comprehensively monitor the operating status of park facilities. The multi-modal perception data includes operating acoustic fingerprint signals, infrared thermal imaging maps, and current harmonic characteristics, which reflect the operating status of the facilities from different angles and can provide more comprehensive information; the operating acoustic fingerprint signal refers to the sound signal generated during the operation of the device collected by a distributed acoustic sensor array, reflecting information such as the mechanical vibration and abnormal noise of the device, used to determine whether there are mechanical failures or wear in the device; the infrared thermal imaging map refers to the surface temperature distribution image of the facility obtained by an infrared thermal imaging device, reflecting the thermal distribution of the device, helping to identify overheated areas, thereby determining whether there are heat dissipation problems or potential failure points. The map has a timestamp for recording the time series of temperature changes.

[0029] The current harmonic feature refers to the harmonic components in the current collected by a non-invasive current sensor. Harmonics are the components in the current with frequencies other than the fundamental frequency, usually caused by electrical faults, aging, or non-linear loads of equipment. The current harmonic feature includes the amplitudes and phase offsets of odd harmonics, reflecting the electrical health status of the equipment. A non-invasive current sensor is installed in the power supply circuit and can measure the current without direct contact with the conductor. It obtains current information by sensing the magnetic field generated by the current and has the characteristics of convenient installation, safety, and reliability. A distributed acoustic sensor array refers to a group of acoustic sensors distributed around the park facilities, used to collect the sound signals during equipment operation. Through the collaborative work of multiple sensors, the sound source can be more accurately located and the sound characteristics can be analyzed.

[0030] Deploy non-invasive current sensors in the power supply circuits of various facilities in the target park to collect current harmonic features; install a distributed acoustic sensor array around the facilities to collect operating voiceprint signals; use an infrared thermal imaging device to obtain the temperature field on the surface of the facilities and generate an infrared thermal imaging map with a timestamp. Preferably, the current harmonic feature refers to the amplitudes and phase offsets of odd harmonics in the current collected in real time by the non-invasive current sensor, reflecting the electrical health status of the equipment. For example, an abnormal increase in the harmonic amplitude may indicate an electrical fault in the equipment. The infrared thermal imaging map refers to the infrared thermal imaging device regularly scanning the surface of the facilities, generating a temperature distribution map, and recording the timestamp. By analyzing the temperature changes, heat dissipation problems or local overheating areas of the equipment can be identified. The operating voiceprint signal refers to the sound signal collected by the distributed acoustic sensor array during equipment operation, and the voiceprint features are extracted through acoustic analysis techniques. For example, abnormal high-frequency noise may indicate mechanical wear of the equipment.

[0031] By deploying multiple sensors to collect multi-modal perception data, it provides comprehensive and accurate information support for subsequent data fusion and fault diagnosis. The operating voiceprint signal, infrared thermal imaging map, and current harmonic feature respectively reflect the operating status of the facilities from the mechanical, thermal, and electrical dimensions, making up for the deficiencies of single-modal monitoring. For example, mechanical faults may cause abnormal operating voiceprint signals, electrical faults may cause changes in current harmonics, and heat dissipation problems will be reflected as local overheating in the infrared thermal imaging map. Through the collaborative monitoring of multi-modal data, potential faults can be more comprehensively identified, improving the accuracy and reliability of the monitoring system.

[0032] Furthermore, collect the multi-modal perception data of each park facility in the target park to generate a fused data matrix through alignment processing. The method of this application includes:

[0033] According to the operation timeline of the park facilities, align the sampling time windows of the operation acoustic fingerprint signal, infrared thermal imaging map, and current harmonic characteristics; based on the aligned sampling time windows, eliminate the phase delay error between the multi-modal data, and map the data into the feature space to generate the fused data matrix.

[0034] Specifically, the operation timeline refers to the time series during the operation of the park facilities, which is used to record and synchronize the acquisition time points of different modal data. The operation timeline is the basis for multi-modal data alignment, ensuring that the data collected by different sensors can correspond in time; the sampling time window refers to the acquisition time range set for each modal data on the operation timeline. Since the acquisition frequencies and methods of the operation acoustic fingerprint signal, infrared thermal imaging map, and current harmonic characteristics may be different, alignment operations need to be performed through the sampling time window for subsequent fusion processing; the phase delay error refers to the deviation in time between different modal data. For example, the operation acoustic fingerprint signal and current harmonic characteristics may be inconsistent in time due to differences in the response time of sensors or data transmission delays, affecting the accuracy of data fusion.

[0035] The feature space is used to map different modal data into a unified and processable dimension. Each dimension in the feature space represents a feature, such as the frequency feature of the operation acoustic fingerprint signal, the temperature gradient feature of the infrared thermal imaging map, or the amplitude feature of the current harmonic. By mapping the data into the feature space, data fusion and analysis can be conveniently performed; integrate the aligned multi-modal data into a matrix structure. Each row of the fused data matrix represents the multi-modal features at a time point, and each column of the fused data matrix represents the features of one modal. The fused data matrix is the basis for subsequent feature extraction and model training.

[0036] According to the operation timeline of the park facilities, align the sampling time windows of the operation acoustic fingerprint signal, infrared thermal imaging map, and current harmonic characteristics. Specifically, select a reference modal (such as the infrared thermal imaging map), and then adjust the sampling time windows of other modal data to be consistent with the time points of the reference modal. For example, if the sampling frequency of the operation acoustic fingerprint signal is 100 Hz, the sampling frequency of the infrared thermal imaging map is 1 Hz, and the sampling frequency of the current harmonic characteristics is 50 Hz, interpolation is required to align the time points; within the aligned sampling time windows, eliminate the phase delay error between different modal data. For example, a time delay estimation algorithm (such as the cross-correlation method) can be used to calculate the time delay between the operation acoustic fingerprint signal and the current harmonic characteristics, and the phase delay error can be eliminated by adjusting the time series of the data.

[0037] Extract features from the aligned multimodal data and map them into a feature space. For example, extract frequency features and energy features from the running voiceprint signal, extract temperature gradients and local temperature change rates from the infrared thermal imaging map, and extract harmonic amplitudes and phase offsets from the current harmonic features. Then, integrate them into a high-dimensional feature space to form a fused data matrix. By aligning the sampling time windows and eliminating phase delay errors, the temporal consistency and accuracy of the multimodal data are ensured. Mapping the data into the feature space further improves the processability and analysis efficiency of the data. The fused data matrix provides a high-quality data foundation for subsequent feature extraction, model training, and fault diagnosis. For example, by eliminating phase delay errors, misjudgments caused by time inconsistency can be avoided. Through the integration of the feature space, the advantages of multimodal data can be fully utilized to improve the sensitivity and accuracy of fault identification.

[0038] Furthermore, based on the operating characteristics of the facility, construct a facility state evolution map. The method of this application includes:

[0039] Extract the spatial correlation features from the fused data matrix; based on the spatial correlation features in the fused data matrix, predict the temporal evolution trend of the facility health index; through the prediction results, determine the fault risk probability and label it on the topological nodes of the facility state evolution map.

[0040] Specifically, the spatial correlation feature refers to the correlation in spatial distribution between different modal data in the fused data matrix. For example, the current harmonic features of a device may be spatially correlated with local temperature changes (infrared thermal imaging map) or mechanical vibrations (running voiceprint signal), reflecting the interaction between different components inside the device. The temporal evolution trend refers to the change trend of the facility health index over time. By analyzing the spatial correlation features in the fused data matrix, the future change situation of the facility health index can be predicted, potential health problems can be discovered in advance, and a basis for preventive maintenance can be provided.

[0041] The fault risk probability refers to the likelihood of a device failing at a certain future time point, which is calculated based on the temporal evolution trend of the facility health index and the current operating state. The determination of the fault risk probability provides a risk warning function for the facility state evolution map. The topological nodes of the facility state evolution map refer to the key positions in the facility state evolution map, which are used to label information such as the health state of the device, abnormal energy consumption, and fault risk. The topological nodes can be specific components of the device or key points of the operating state. Labeling the fault risk probability on the topological nodes of the facility state evolution map visually displays the health state and potential risks of the device.

[0042] Extract the spatial correlation features between different modality data from the fused data matrix. For example, by analyzing the correlation between current harmonic features and local temperature changes in the infrared thermal imaging map, or the correlation between running acoustic signals and the surface temperature of the device; based on the extracted spatial correlation features, use time series analysis methods (such as ARIMA model, long short-term memory network LSTM) to predict the future change trend of the facility health index. For example, if the amplitude of the current harmonic of the device is positively correlated with the local temperature change and the correlation increases over time, it indicates a decline in the health index.

[0043] According to the time evolution trend of the health index, combined with the current energy consumption anomaly coefficient and other operating state indicators, calculate the failure risk probability. For example, if the health index shows a downward trend and the energy consumption anomaly coefficient is high, the failure risk probability increases. Mark the calculated failure risk probability on the topological nodes of the facility state evolution map; by extracting the spatial correlation features, the internal relationship between different modality data can be deeply understood, so as to more accurately evaluate the health state of the device. Predicting the time evolution trend of the facility health index provides an early warning for preventive maintenance and avoids sudden failures. The determination and marking of the failure risk probability further enhance the function of the facility state evolution map, enabling it to intuitively display the health state and potential risks of the device, and providing strong support for the intelligence and automation of the entire monitoring system.

[0044] Furthermore, according to the facility health index and energy consumption anomaly coefficient in the facility state evolution map, configure the first monitoring period, and identify a class of potential failure modes under the initial failure identification rule. The method of the present application further includes:

[0045] Configure the first threshold range corresponding to the facility health index and energy consumption anomaly coefficient, and the first threshold range is adaptively adjusted according to the environmental temperature and load rate; in the initial failure identification rule, use the first threshold range and the first monitoring period to establish the first state evaluation decision chain corresponding to a class of potential failure modes, and monitor the facility operation data in real time; use the first state evaluation decision chain to perform multi-modal data backtracking analysis, and extract multiple abnormal feature segments of the occurrence of the failure; according to the multiple abnormal feature segments, perform state transition feedback optimization on the first state evaluation decision chain.

[0046] Specifically, the first threshold range refers to the initial threshold interval set for the facility health index and the energy consumption anomaly coefficient, which is used to determine whether the facility operation status is normal. This threshold range will be dynamically adjusted according to the environmental temperature and the load rate to adapt to the operation status under different working conditions. Adaptive adjustment means that the threshold range will change dynamically according to the real-time environmental conditions (such as temperature, load rate, etc.). For example, when the environmental temperature rises, the threshold of the energy consumption anomaly coefficient may be relaxed accordingly because the energy consumption of the equipment may naturally increase when operating in a high-temperature environment.

[0047] The first state assessment decision is used to evaluate the operation status of the facility according to the first threshold range and the first monitoring period. Multi-modal data backtracking analysis means that after detecting a potential fault, the historical multi-modal data is backtracked and analyzed to extract abnormal feature segments before and after the fault occurs, including abnormal noises in the running voiceprint signal, temperature abnormal areas in the infrared thermal imaging map, or mutations in the current harmonic characteristics, etc. State transfer feedback optimization means that according to the results of the multi-modal data backtracking analysis, the first state assessment decision chain is optimized by adjusting the decision logic or the threshold range to enable it to more accurately identify the fault mode and reduce false alarms or missed alarms.

[0048] According to the normal operation status of the facility, the initial threshold range of the health index and the energy consumption anomaly coefficient is set. At the same time, the environmental temperature and the load rate are introduced as dynamic adjustment factors to enable the threshold range to adaptively change according to the real-time working conditions. For example, when the load rate increases, the threshold of the energy consumption anomaly coefficient may be appropriately increased to avoid misjudging as abnormal due to the increase in normal load. Combining the first threshold range and the first monitoring period, a decision logic for real-time monitoring of the facility operation status is constructed. For example, if the facility health index is lower than the set threshold or the energy consumption anomaly coefficient exceeds the threshold, an alarm for a type of potential fault mode is triggered.

[0049] When a type of potential fault mode is identified, the historical multi-modal data is backtracked and analyzed to extract multiple abnormal feature segments before and after the fault occurs. According to the results of the backtracking analysis, the first state assessment decision chain is optimized. For example, if it is found that a certain abnormal feature segment frequently appears in multiple faults, it can be added to the decision logic to enhance the accuracy of the decision chain; or the threshold range can be adjusted according to the actual fault situation to make it more in line with the actual operation status. By adaptively adjusting the threshold range, it can more flexibly cope with the operation status under different working conditions and reduce false alarms or missed alarms caused by environmental changes. The first state assessment decision chain provides a clear logical basis for real-time monitoring and can quickly identify a type of potential fault mode. Multi-modal data backtracking analysis and state transfer feedback optimization further improve the intelligence level of the system, enabling it to dynamically adjust the decision logic according to the actual operation data and improve the accuracy and reliability of fault identification.

[0050] Furthermore, by performing pattern matching on the first monitoring cycle and a type of potential fault mode, the second monitoring cycle and a second type of potential fault mode with the historical fault database, the method of the present application includes:

[0051] Configure the second threshold range corresponding to the facility health index and the energy consumption anomaly coefficient; in the initial fault identification rule, use the second threshold range and the second monitoring cycle to establish a second-state evaluation decision chain corresponding to the second type of potential fault mode; use the first state evaluation decision chain and the second-state evaluation decision chain to configure a bidirectional traversal architecture, which is used to perform cross-verification with the first type of potential fault mode and the second type of potential fault mode.

[0052] Specifically, similar to the first threshold range, the second threshold range is another set of threshold intervals set for the facility health index and the energy consumption anomaly coefficient, specifically for identifying the second type of potential fault mode. The second threshold range may be adjusted according to different fault characteristics or more stringent monitoring requirements to adapt to the identification of the second type of potential fault mode; the second-state evaluation decision chain is a decision logic constructed based on the second threshold range and the second monitoring cycle, used to evaluate whether there is a second type of potential fault mode in the facility, similar to the first state evaluation decision chain.

[0053] The bidirectional traversal architecture is a system architecture for cross-verifying the first and second types of potential fault modes. By simultaneously running the first state evaluation decision chain and the second-state evaluation decision chain, and performing information interaction and verification between the two, it ensures the accuracy and comprehensiveness of fault identification. The bidirectional traversal architecture reduces the possibility of false alarms and missed alarms by comparing the identification results of the two types of fault modes; according to the characteristics of the second type of potential fault mode, set the second threshold range for the facility health index and the energy consumption anomaly coefficient; combine the second threshold range and the second monitoring cycle to construct a decision logic for identifying the second type of potential fault mode. For example, if the energy consumption anomaly coefficient of the facility exceeds the second threshold range and the fault risk probability reaches a certain level, an alarm for the second type of potential fault mode is triggered.

[0054] Integrate the first - state evaluation decision chain and the second - state evaluation decision chain into a bidirectional traversal architecture. The bidirectional traversal architecture ensures the accuracy and comprehensiveness of fault identification by cross - validating the identification results of type - one and type - two potential fault modes. For example, if the first - state evaluation decision chain identifies a type - one potential fault mode while the second - state evaluation decision chain does not identify the corresponding type - two potential fault mode, further data analysis is required to determine whether there are false alarms or missed detections. By configuring the second threshold range and the second - state evaluation decision chain, more accurate monitoring and identification of type - two potential fault modes can be achieved. The introduction of the bidirectional traversal architecture further improves the reliability of the system. By cross - validating the identification results of type - one and type - two potential fault modes, the problem of false alarms or missed detections that may be caused by a single decision chain is reduced. The dual - verification mechanism ensures the accuracy of fault identification and enhances the accuracy and comprehensiveness of fault identification.

[0055] Furthermore, the bidirectional traversal architecture is used to cross - validate with type - one potential fault modes and type - two potential fault modes. The method of the present application includes:

[0056] When the type - one potential fault mode and the type - two potential fault mode are triggered simultaneously, activate the priority - ranking mechanism of the bidirectional traversal architecture; under the priority - ranking mechanism, add virtual aggregation index items; use the virtual aggregation index items to perform pattern matching in the historical fault database.

[0057] Specifically, in the bidirectional traversal architecture, when the type - one potential fault mode and the type - two potential fault mode are triggered simultaneously, it is necessary to perform a priority ranking on the two fault modes. The priority - ranking mechanism is used to determine which fault mode is more urgent or critical, so as to be processed first, usually based on factors such as the severity of the fault, the risk probability, and the impact scope to determine the priority; the virtual aggregation index item refers to a temporary index item introduced during the pattern - matching process to improve the matching efficiency and accuracy. The virtual aggregation index item is dynamically generated according to the characteristics of the currently identified fault mode and is used to quickly locate the historical records similar to the current fault mode in the historical fault database, which can accelerate the pattern - matching process and improve the accuracy of the matching; pattern matching refers to comparing and analyzing the currently identified fault mode (including type - one and type - two potential fault modes) with the records in the historical fault database to determine whether there are similar fault modes.

[0058] When a type-I potential failure mode and a type-II potential failure mode are triggered simultaneously, the bidirectional traversal architecture will activate the priority sorting mechanism. The system will sort the two failure modes according to preset rules (such as failure risk probability, health index, energy consumption anomaly coefficient, etc.) to determine which failure mode is more urgent or critical. For example, if the type-I potential failure mode has a higher risk probability, it will be given a higher priority; under the priority sorting mechanism, virtual aggregation index items are dynamically generated based on the characteristics of the current failure mode, which are generated based on the key characteristics of the failure mode (such as a specific health index range, energy consumption anomaly coefficient, failure risk probability, etc.) and are used to quickly locate similar failure records in the historical failure database.

[0059] Using the virtual aggregation index items, pattern matching is performed in the historical failure database. The system will quickly retrieve historical records similar to the current failure mode according to the virtual aggregation index items and calculate the similarity. For example, by calculating the feature vector distance (such as Euclidean distance or cosine similarity) between the current failure mode and the historical records, it is determined whether there is a matching historical failure mode, and the matching result will be used to verify the accuracy of the current failure mode; when the type-I and type-II potential failure modes are triggered simultaneously, the priority sorting mechanism can quickly determine which failure mode is more urgent, so as to reasonably allocate resources and prioritize the handling of key issues. The introduction of virtual aggregation index items significantly improves the efficiency and accuracy of pattern matching, enabling the system to quickly find records similar to the current failure mode in a large amount of historical failure data, using the pattern matching mechanism to verify the accuracy of the current failure mode, providing strong support for subsequent fault diagnosis and optimization, and further improving the intelligent level and reliability of the system.

[0060] Furthermore, using the virtual aggregation index items for pattern matching in the historical failure database, the method of the present application includes:

[0061] Using the virtual aggregation index items, through correlation calculation, set the first feature pointer corresponding to the type-I potential failure mode and the second feature pointer corresponding to the type-II potential failure mode; through the first feature pointer corresponding to the type-I potential failure mode, search for the first failure mode matching set; through the second feature pointer corresponding to the type-II potential failure mode, search for the second failure mode matching set; map the first failure mode matching set and the second failure mode matching set to the state transition space to generate a state transition trajectory.

[0062] Specifically, a virtual aggregated index item refers to an index item dynamically generated based on the key features of the current fault mode during the pattern matching process, which is used to quickly locate records similar to the current fault mode in the historical fault database and improve the matching efficiency; correlation calculation refers to evaluating the similarity between the current fault mode and historical fault data through mathematical methods (such as similarity calculation, distance metric, etc.). The higher the correlation, the greater the similarity between the two, usually based on the distance between feature vectors (such as Euclidean distance, cosine similarity, etc.); the first feature pointer / second feature pointer refers to an index in the historical fault database generated according to the correlation calculation result, pointing to a specific fault mode. The first feature pointer corresponds to a class of potential fault modes, and the second feature pointer corresponds to a second class of potential fault modes, which are used to quickly locate the historical record most similar to the current fault mode.

[0063] The fault mode matching set refers to a set of historical records similar to the current fault mode searched in the historical fault database through feature pointers. The first fault mode matching set corresponds to a class of potential fault modes, and the second fault mode matching set corresponds to a second class of potential fault modes; the state transition space is a mathematical model or data structure used to describe the conversion relationship between fault modes, which records the state changes of fault modes at different time points and reflects the development trend and evolution process of faults; the state transition trajectory refers to the evolution path of the fault mode from the initial state to the current state in the state transition space. By analyzing the state transition trajectory, the future development trend of faults can be predicted, providing a basis for fault diagnosis and optimization.

[0064] Use virtual aggregated index items to evaluate the similarity between a class of potential fault modes and a second class of potential fault modes and historical fault data through correlation calculation (such as calculating the similarity or distance between feature vectors); according to the correlation calculation result, set the first feature pointer and the second feature pointer for a class of potential fault modes and a second class of potential fault modes respectively, pointing to the historical record most similar to the current fault mode in the historical fault database; through the first feature pointer, search for historical records similar to a class of potential fault modes in the historical fault database to generate the first fault mode matching set; through the second feature pointer, search for historical records similar to a second class of potential fault modes in the historical fault database to generate the second fault mode matching set.

[0065] Map the first failure mode matching set and the second failure mode matching set to the state transition space, which records the conversion relationship and evolution process between failure modes; in the state transition space, generate the state transition trajectory of the failure mode according to the historical records in the matching set, and the state transition trajectory reflects the evolution path of the failure mode from the initial state to the current state, providing a basis for fault diagnosis and optimization; through correlation calculation and feature pointer setting, the system can quickly locate the historical record most similar to the current failure mode, improving the accuracy and efficiency of fault diagnosis. The generation of the state transition trajectory provides support for the dynamic analysis of faults. By analyzing the evolution path of the failure mode, the future development trend of the fault can be predicted, preventive measures can be taken in advance, and the impact of the fault on the operation of park facilities can be reduced.

[0066] Furthermore, the method of this application includes:

[0067] Set the first feature pointer corresponding to the first type of potential failure mode through correlation calculation ; where is the first feature pointer, D is the historical fault database, represents the feature vector corresponding to the first failure mode matching set, is the feature vector corresponding to the first type of potential failure mode, is used to characterize the standard deviation of the Gaussian kernel associated with the first failure mode matching set, and is used to control the attenuation speed of similarity.

[0068] Specifically, is the first feature pointer, indicating the index or position of the failure mode in the historical fault database D that is most similar to the first type of potential failure mode ; Search in the historical fault database D to find the that maximizes the value of ; is used to calculate the similarity between the current potential failure mode and the historical failure mode ; where is the feature vector corresponding to the first type of potential failure mode, is a failure mode feature vector in the historical fault database D, refers to the Euclidean norm, indicating and the distance between refers to the standard deviation of the Gaussian kernel used to characterize the association of the first failure mode matching set. When using the Gaussian kernel for similarity calculation, it is used to control the attenuation speed of similarity. The larger the standard deviation, the slower the attenuation speed of similarity, that is, the more gentle the decrease in similarity. Conversely, the smaller the standard deviation, the faster the attenuation speed of similarity, that is, the steeper the decrease in similarity.

[0069] Using the feature vectors of the current type of potential fault mode, compare them with the feature vectors in the historical fault database. Further, through correlation calculation, set the first feature pointer corresponding to the type of potential fault mode ; Through correlation calculation and the setting of the first feature pointer, quickly locate the historical record most similar to the current fault mode, improving the accuracy and efficiency of fault diagnosis. The introduction of the Gaussian kernel standard deviation makes the similarity calculation more flexible, and can adjust the attenuation speed of the similarity according to actual needs, thus more accurately reflecting the similarity between fault modes.

[0070] In summary, the beneficial effects of the embodiments of the present application are as follows:

[0071] Since multi-modal perception data of each park facility in the target park is collected, a fusion data matrix is generated through alignment processing and input into a pre-trained adaptive association model to extract facility operation features and construct a facility state evolution map; according to the facility health index and energy consumption anomaly coefficient, and the fault risk probability and energy consumption anomaly coefficient in the facility state evolution map, the first monitoring period and the second monitoring period are respectively configured, and a type of potential fault mode and a second type of potential fault mode are identified under the initial fault identification rule, and pattern matching is performed with the historical fault database, and the monitoring period and fault identification rule of each park facility are optimized according to the matching result, and a fault monitoring report is output. The present application provides a method and system for monitoring the operation state of park facilities based on multi-modal perception, constructs a facility state evolution map, combines the time prediction of the health index and the spatial annotation of the fault risk probability, and identifies hidden faults in advance. At the same time, through multi-modal data retrospective analysis and state transfer feedback optimization, the sensitivity of fault feature extraction is improved, and incremental learning and model iteration are used to continuously optimize the adaptive association model, improving the accuracy and comprehensiveness of fault identification.

[0072] Embodiment 2

[0073] Based on the same inventive concept as the method for monitoring the operation state of park facilities based on multi-modal perception in the foregoing embodiment, as Figure 2 shown, the embodiments of the present application provide a system for monitoring the operation state of park facilities based on multi-modal perception, wherein the system includes:

[0074] A data acquisition module M100, configured to collect multi-modal perception data of each park facility in the target park, generate a fusion data matrix through alignment processing, and input it into a pre-trained adaptive association model to extract facility operation features.

[0075] A map construction module M200, configured to construct a facility state evolution map based on the facility operation features, including a facility health index, an energy consumption anomaly coefficient, and a fault risk probability.

[0076] The first identification module M300 is configured to configure a first monitoring period according to the facility health index and the energy consumption anomaly coefficient in the facility status evolution map, and identify a first type of potential fault mode under the initial fault identification rule.

[0077] The second identification module M400 is configured to configure a second monitoring period according to the fault risk probability and the energy consumption anomaly coefficient in the facility status evolution map, and identify a second type of potential fault mode under the initial fault identification rule.

[0078] The pattern matching module M500 is configured to perform pattern matching on the first monitoring period and the first type of potential fault mode, the second monitoring period and the second type of potential fault mode with the historical fault database, optimize the monitoring period and the fault identification rule of each park facility according to the matching result, and output a fault monitoring report.

[0079] Furthermore, the data acquisition module M100 is further configured to execute the following method:

[0080] The multi-modal perception data includes operating voiceprint signals, infrared thermal imaging maps, and current harmonic characteristics.

[0081] In the power supply circuits corresponding to each park facility in the target park, non-intrusive current sensors are deployed to obtain current harmonic characteristics including odd harmonic amplitudes and phase offsets.

[0082] Obtain the surface temperature field of the park facility and generate an infrared thermal imaging map with a time stamp.

[0083] Use a distributed acoustic sensor array to collect operating voiceprint signals.

[0084] Furthermore, the data acquisition module M100 is further configured to execute the following method:

[0085] Align the sampling time windows of the operating voiceprint signals, infrared thermal imaging maps, and current harmonic characteristics according to the operation time axis of the park facility.

[0086] Based on the aligned sampling time windows, eliminate the phase delay error between the multi-modal data, and map the data into the feature space to generate the fusion data matrix.

[0087] Furthermore, the map construction module M200 is configured to execute the following method:

[0088] Extract the spatial correlation features in the fusion data matrix.

[0089] Based on the spatial correlation features in the fusion data matrix, predict the time evolution trend of the facility health index.

[0090] Determine the fault risk probability based on the prediction results and label it on the topological nodes of the facility status evolution map.

[0091] Furthermore, the first identification module M300 is also used to execute the following method:

[0092] Configure the first threshold ranges corresponding to the facility health index and the energy consumption anomaly coefficient, and the first threshold ranges are adaptively adjusted according to the environmental temperature and the load rate.

[0093] In the initial fault identification rule, use the first threshold ranges and the first monitoring period to establish the first state evaluation decision chain corresponding to a class of potential fault modes, and monitor the facility operation data in real time.

[0094] Use the first state evaluation decision chain to perform multi-modal data backtracking analysis and extract multiple abnormal feature segments where the fault occurs.

[0095] According to the multiple abnormal feature segments, perform state transition feedback optimization on the first state evaluation decision chain.

[0096] Furthermore, the pattern matching module M500 is also used to execute the following method:

[0097] Configure the second threshold ranges corresponding to the facility health index and the energy consumption anomaly coefficient.

[0098] In the initial fault identification rule, use the second threshold ranges and the second monitoring period to establish the second state evaluation decision chain corresponding to a second class of potential fault modes.

[0099] Use the first state evaluation decision chain and the second state evaluation decision chain to configure a bidirectional traversal architecture, and the bidirectional traversal architecture is used for cross-verification with a class of potential fault modes and a second class of potential fault modes.

[0100] Furthermore, the pattern matching module M500 is also used to execute the following method:

[0101] When the class of potential fault modes and the second class of potential fault modes are triggered simultaneously, activate the priority sorting mechanism of the bidirectional traversal architecture.

[0102] Under the priority sorting mechanism, add virtual aggregation index items.

[0103] Use the virtual aggregation index items to perform pattern matching in the historical fault database.

[0104] Furthermore, the pattern matching module M500 is also used to execute the following method:

[0105] Using the virtual aggregated index items, the first feature pointer corresponding to the first type of potential failure mode and the second feature pointer corresponding to the second type of potential failure mode are set through correlation calculation.

[0106] Search the first failure mode matching set through the first feature pointer corresponding to the first type of potential failure mode; search the second failure mode matching set through the second feature pointer corresponding to the second type of potential failure mode.

[0107] Map the first failure mode matching set and the second failure mode matching set to the state transition space to generate a state transition trajectory.

[0108] Further, the pattern matching module M500 is further configured to execute the following method:

[0109] Set the first feature pointer corresponding to the first type of potential failure mode through correlation calculation .

[0110] Wherein, is the first feature pointer, D is the historical failure database, represents the feature vector corresponding to the first failure mode matching set, is the feature vector corresponding to the first type of potential failure mode, is used to characterize the Gaussian kernel standard deviation associated with the first failure mode matching set and is used to control the attenuation rate of similarity.

[0111] In summary, any step can be stored as a computer instruction or program in an unrestricted computer memory and can be called and recognized by an unrestricted computer processor, and no redundant restrictions are imposed here.

[0112] Furthermore, the above technical solutions only represent the preferred technical solutions of the technical solutions of the embodiments of the present application. Some changes that those skilled in the art may make to some parts thereof all reflect the principles of the novel embodiments of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application.

Claims

1. A method for monitoring the operating status of park facilities based on multi-modal perception, characterized in that, The method includes: Collecting multi-modal perception data of each park facility in the target park, generating a fusion data matrix through alignment processing, and inputting it into a pre-trained adaptive association model to extract facility operation characteristics; Constructing a facility state evolution map based on the facility operation characteristics, including a facility health index, an energy consumption anomaly coefficient, and a fault risk probability; Configuring a first monitoring period according to the facility health index and the energy consumption anomaly coefficient in the facility state evolution map, and identifying a first type of potential fault mode under the initial fault identification rule; Configuring a second monitoring period according to the fault risk probability and the energy consumption anomaly coefficient in the facility state evolution map, and identifying a second type of potential fault mode under the initial fault identification rule; Performing pattern matching on the first monitoring period and the first type of potential fault mode, the second monitoring period and the second type of potential fault mode with a historical fault database, optimizing the monitoring period and the fault identification rule of each park facility with the matching result, and outputting a fault monitoring report; Among them, configuring a first monitoring period according to the facility health index and the energy consumption anomaly coefficient in the facility state evolution map, and identifying a first type of potential fault mode under the initial fault identification rule includes: Configuring a first threshold range corresponding to the facility health index and the energy consumption anomaly coefficient, and the first threshold range is adaptively adjusted according to the environmental temperature and the load rate; In the initial fault identification rule, using the first threshold range and the first monitoring period, establishing a first state evaluation decision chain corresponding to the first type of potential fault mode, and performing real-time monitoring on the facility operation data; Using the first state evaluation decision chain to perform multi-modal data backtracking analysis and extract multiple abnormal feature segments where the fault occurs; Optimizing the state transfer feedback of the first state evaluation decision chain according to the multiple abnormal feature segments; Among them, performing pattern matching on the first monitoring period and the first type of potential fault mode, the second monitoring period and the second type of potential fault mode with a historical fault database includes: Configuring a second threshold range corresponding to the facility health index and the energy consumption anomaly coefficient; In the initial fault identification rule, using the second threshold range and the second monitoring period, establishing a second state evaluation decision chain corresponding to the second type of potential fault mode; Using the first state evaluation decision chain and the second state evaluation decision chain to configure a bidirectional traversal architecture, and the bidirectional traversal architecture is used to perform cross-validation with the first type of potential fault mode and the second type of potential fault mode; The bidirectional traversal architecture is used to perform cross-validation with the first type of potential fault mode and the second type of potential fault mode, including: When the first type of potential fault mode and the second type of potential fault mode are triggered simultaneously, activating the priority sorting mechanism of the bidirectional traversal architecture; Adding a virtual aggregation index item under the priority sorting mechanism; Using the virtual aggregation index item to perform pattern matching in the historical fault database.

2. The method for monitoring the operation status of park facilities based on multi-modal perception according to claim 1, wherein, The multi-modal perception data includes an operating voiceprint signal, an infrared thermal imaging map, and a current harmonic feature; Deploying a non-intrusive current sensor in the power supply circuit corresponding to each park facility in the target park to obtain a current harmonic feature including the odd harmonic amplitude and the phase offset. Obtain the surface temperature field of the park facilities and generate an infrared thermal imaging map with a timestamp; Use a distributed acoustic sensor array to collect operating acoustic fingerprint signals.

3. The method for monitoring the operation status of park facilities based on multi-modal perception according to claim 2, wherein, Collect multi-modal perception data of each park facility in the target park, and generate a fused data matrix through alignment processing. The method includes: According to the operating time axis of the park facilities, align the sampling time windows of the operating acoustic fingerprint signals, infrared thermal imaging maps, and current harmonic characteristics; Based on the aligned sampling time windows, eliminate the phase delay error between multi-modal data and map the data into the feature space to generate the fused data matrix.

4. The method for monitoring the operation status of park facilities based on multimodal perception according to claim 1, characterized in that, Based on the operating characteristics of the facilities, construct a facility state evolution map. The method includes: Extract the spatial correlation features from the fused data matrix; Based on the spatial correlation features in the fused data matrix, predict the time evolution trend of the facility health index; Through the prediction results, determine the failure risk probability and label it on the topological nodes of the facility state evolution map.

5. The method for monitoring the operation status of park facilities based on multi-modal perception according to claim 1, wherein Use the virtual aggregation index item to perform pattern matching in the historical failure database. The method includes: Use the virtual aggregation index item to set the first feature pointer corresponding to the first type of potential failure mode and the second feature pointer corresponding to the second type of potential failure mode through correlation calculation; Through the first feature pointer corresponding to the first type of potential failure mode, search for the first failure mode matching set; through the second feature pointer corresponding to the second type of potential failure mode, search for the second failure mode matching set; Map the first failure mode matching set and the second failure mode matching set to the state transition space to generate a state transition trajectory.

6. The method for monitoring the operation status of park facilities based on multimodal perception according to claim 5, wherein Set the first feature pointer corresponding to the first type of potential failure mode through relevance calculation ; Among them, is the first feature pointer, D is the historical fault database, represents the feature vector corresponding to the first fault mode matching set, is the feature vector corresponding to a type of potential fault mode, is used to characterize the Gaussian kernel standard deviation associated with the first fault mode matching set and control the attenuation rate of similarity.

7. A monitoring system for the operating status of park facilities based on multimodal perception, characterized in that, For implementing the method for monitoring the operating state of park facilities based on multi-modal perception according to any one of claims 1-6, the system includes: A data acquisition module for collecting multi-modal perception data of each park facility in the target park, generating a fused data matrix through alignment processing, and inputting it into a pre-trained adaptive correlation model to extract facility operating characteristics; A map construction module for constructing a facility state evolution map based on the facility operating characteristics, including a facility health index, an energy consumption anomaly coefficient, and a failure risk probability; A first identification module for configuring a first monitoring period according to the facility health index and energy consumption anomaly coefficient in the facility state evolution map, and identifying the first type of potential failure mode under the initial failure identification rule; A second identification module for configuring a second monitoring period according to the failure risk probability and energy consumption anomaly coefficient in the facility state evolution map, and identifying the second type of potential failure mode under the initial failure identification rule; A pattern matching module for performing pattern matching on the first monitoring period and the first type of potential failure mode, the second monitoring period and the second type of potential failure mode with the historical failure database, optimizing the monitoring period and failure identification rule of each park facility with the matching results, and outputting a failure monitoring report.

Images