Intelligent park management method and system based on multi-source data fusion

Abstract

The invention discloses a park intelligent management method and system based on multi-source data fusion, and relates to the technical field of data intelligent management. Comprising the steps of collecting data to form an operation feature data set; key indexes are extracted, and key index data are mapped to an equipment operation state space; constructing an equipment life prediction model based on the extracted key index data; analyzing the real-time operation feature data of the key equipment, generating a maintenance suggestion according to a health assessment result and the distribution of real-time feature points in a state space, and marking potential maintenance risks; and making an equipment maintenance plan according to the equipment health assessment result and the life prediction value. According to the method, the limitation of neglecting multi-dimensional characteristics and dynamic changes of the equipment in a traditional method is avoided, so that the equipment failure rate and the downtime are effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent data management, and in particular to an intelligent park management method and system based on multi-source data fusion. Background Art

[0002] With the continuous development of industrial automation and intelligent technology, intelligent park management has become an increasingly important research direction in terms of improving operational efficiency, reducing equipment failure rates, and optimizing resource allocation. Traditional park equipment management relies on regular manual inspections and experience-based maintenance strategies. This management method is not only inefficient, but also prone to missed detection of potential equipment failures, which in turn leads to increased equipment downtime and reduced production efficiency.

[0003] Although the existing technology has made some progress in intelligent equipment management, there are still some significant shortcomings. First, most existing intelligent management methods still rely on a single monitoring data source, such as only using vibration or temperature data to evaluate equipment health. This method may ignore the complex interactive relationships between equipment in various operating environments. Secondly, data analysis in existing technologies is often limited to predictions based on static models, and fails to effectively utilize dynamic data and changes in the equipment operating environment, resulting in inaccurate life prediction results and difficulty in adapting to the actual situation of performance changes of equipment during long-term operation. In addition, many prediction models ignore the fusion of multi-source data and are unable to comprehensively consider multiple characteristics of the equipment such as vibration, temperature and load, resulting in constraints on prediction accuracy and the effectiveness of maintenance decisions. Summary of the invention

[0004] In view of the problems existing in the above-mentioned background technology, the present invention proposes a park intelligent management method and system based on multi-source data fusion.

[0005] Therefore, the problem to be solved by the present invention is how to solve the problems of low prediction accuracy and poor dynamic adaptability that are commonly found in equipment health management in a park.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, the present invention provides a park intelligent management method based on multi-source data fusion, which includes, S1: collecting vibration data, temperature data and load data of key equipment, performing timestamp synchronization processing on the collected data, and normalizing and noise filtering processing according to the key equipment operation standards to form an operation feature data set; S2: using sliding window technology and frequency domain analysis algorithm to extract key indicators from the standardized operation feature data set, and map the key indicator data to the equipment operation state space; S3: based on the extracted key indicator data, multidimensional regression analysis and historical life data are used for training to build an equipment life prediction model; S4: using the life prediction model to analyze the real-time operation feature data of key equipment, and generate maintenance suggestions and mark potential maintenance risks according to the health assessment results and the distribution of real-time feature points in the state space; S5: according to the equipment health assessment results and life prediction values, combined with the equipment operation schedule and resource allocation of the park, formulate an equipment maintenance plan, and dynamically adjust the plan parameters according to the maintenance execution feedback data to reduce equipment downtime.

[0008] As a preferred solution of the park intelligent management method based on multi-source data fusion described in the present invention, wherein: S2 includes the following steps: applying sliding window segmentation to the operation feature data set according to the time series, defining each window as a fixed length, and setting the window sliding step to form multiple continuous time period subsets; using fast Fourier transform to calculate the main frequency distribution in the window, and extracting the first three components with the largest frequency amplitude as vibration features, denoted as F vib ={f1,f2,f3}, where f i is the i-th main frequency component of the vibration signal in the window; the temperature fluctuation amplitude ΔT in the statistical window; the maximum load value P is captured through the data peak recognition algorithm in the sliding window max and peak frequency; construct a multidimensional feature vector [F vib ,ΔT,P max ] describes the comprehensive operating status of the equipment in the window; uses a dimensionality reduction technique based on principal component analysis to map the multidimensional feature vector to a two-dimensional plane in the equipment operating status space, defines the position of the status point as [x, y], and stores the status point sequence in chronological order to the feature database, and marks the associated index with the original time series.

[0009] As a preferred solution of the park intelligent management method based on multi-source data fusion described in the present invention, the multi-dimensional feature vector [F vib ,ΔT,P max ] is mapped to the two-dimensional coordinates of the running state space, and the operation is as follows:

[0010]

[0011] Where [x,y] T is the two-dimensional coordinate of the running state space, and W is the state mapping weight matrix.

[0012] As a preferred solution of the park intelligent management method based on multi-source data fusion described in the present invention, wherein: S3 includes the following steps: based on the generated two-dimensional plane data of the operating state space, extracting key feature points in the state space; combining the historical operation data of the equipment, using a multidimensional regression analysis method to train the extracted two-dimensional plane feature points to construct a life prediction model; after the life prediction model training is completed, dynamically adjusting the feature weight parameters in the life prediction model to adapt to changes in the equipment operating state.

[0013] As a preferred solution of the park intelligent management method based on multi-source data fusion described in the present invention, the construction of the life prediction model includes: realizing the prediction calculation of equipment life by analyzing the corresponding relationship between the distribution law of feature points and equipment life data. The specific formula is as follows:

[0014]

[0015] Where L is the predicted value of the remaining life of the equipment, is a nonlinear mapping function, N is the number of historical feature points, α i is the regression weight of the feature point, β is the regression weight of the historical life data, H is the historical life data of the equipment, P z =[x s ,y s ] T is the real-time running state coordinate; the calculation of the distribution of the real-time feature points in the state space, that is, the calculation of the Euclidean distance d between the real-time feature points and the historical health area real .

[0016] As a preferred solution of the park intelligent management method based on multi-source data fusion described in the present invention, the calculation of the health assessment is as follows:

[0017]

[0018] Among them, Health is the health index of the device, L y It is the design reference life value of the equipment.

[0019] As a preferred solution of the park intelligent management method based on multi-source data fusion described in the present invention, the generating maintenance suggestion includes: if the distance d real Exceeding the set radius R of the health zone he, it means that the current state deviates from the healthy area; according to the two trigger conditions, a comprehensive maintenance risk mark is generated, including: if Health <H th , and d real >R he , then comprehensive maintenance is recommended; if Health <H th , life-related maintenance is recommended; if d real >R he , it is recommended to perform a state deviation check; where H th is the health threshold.

[0020] In a second aspect, the present invention provides a park intelligent management system based on multi-source data fusion, which includes:

[0021] The acquisition module is used to collect vibration data, temperature data and load data of key equipment, synchronize the collected data with the timestamp, and perform normalization and noise filtering according to the key equipment operation standards to form an operation feature data set;

[0022] An extraction module is used to extract key indicators from the standardized operation feature data set using sliding window technology and frequency domain analysis algorithm, and map the key indicator data to the equipment operation state space;

[0023] A module construction module is used to construct an equipment life prediction model based on the extracted key indicator data by using multidimensional regression analysis and historical life data for training;

[0024] An evaluation module, used to analyze the real-time operation characteristic data of key equipment using the life prediction model, generate maintenance suggestions and mark potential maintenance risks based on the health assessment results and the distribution of real-time characteristic points in the state space;

[0025] The maintenance module is used to formulate equipment maintenance plans based on equipment health assessment results and life prediction values, combined with the equipment operation schedule and resource allocation of the park, and dynamically adjust plan parameters based on maintenance execution feedback data to reduce equipment downtime.

[0026] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the campus intelligent management method based on multi-source data fusion as described in the first aspect of the present invention are implemented.

[0027] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the campus intelligent management method based on multi-source data fusion as described in the first aspect of the present invention are implemented.

[0028] The beneficial effects of the present invention are as follows: the present invention collects and integrates the vibration data, temperature data and load data of the equipment, uses sliding window technology and frequency domain analysis algorithm to extract key indicators, and trains the life prediction model based on historical life data and multidimensional regression analysis. This comprehensive analysis of multidimensional data can better reflect the health status of the equipment under various operating conditions. At the same time, combined with the operating environment of the equipment and the dynamic update of real-time data, our invention can achieve more accurate equipment life prediction, and generate real-time health assessment and maintenance recommendations based on this, avoiding the limitations of traditional methods that ignore the multidimensional characteristics and dynamic changes of equipment, thereby effectively reducing equipment failure rate and downtime. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0030] Figure 1 This is a flow chart of the park intelligent management method based on multi-source data fusion.

[0031] Figure 2 This is a structural diagram of the park intelligent management system based on multi-source data fusion. DETAILED DESCRIPTION

[0032] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0033] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0034] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0035] Example 1

[0036] Reference Figure 1-2 , which is the first embodiment of the present invention, and provides a park intelligent management method based on multi-source data fusion, including:

[0037] S1: Vibration data, temperature data and load data of key equipment are collected through multi-source sensing devices deployed on key equipment in the park. The collected data are synchronized with timestamps and normalized and noise filtered according to the key equipment operation standards to form a usable operation feature data set.

[0038] Specifically, multi-source sensors are installed at major locations of key equipment, including vibration sensors, temperature sensors, and load sensors. Vibration sensors are installed on bearings or rotating parts of the equipment to monitor vibration frequency and amplitude. Temperature sensors are arranged in heat source areas or heat dissipation locations of the equipment to monitor changes in the operating temperature of the equipment. Load sensors are integrated in transmission or force-bearing parts of the equipment to capture load fluctuations during operation of the equipment.

[0039] Configure the sampling frequency of the sensor (such as 10kHz for vibration signal, 1Hz for temperature signal, and 100Hz for load signal) to ensure that the signal capture accuracy requirements are met.

[0040] The sensor data stream is received in real time through the edge computing device, and the collection time of each data point is marked with a timestamp.

[0041] Furthermore, after collecting data from different sensors, they are aligned by timestamp to unify different types of sensor data into a unified time dimension, including: using a linear interpolation algorithm to fill some timestamp gaps to compensate for the discontinuity of data collection.

[0042] For sensor data with significant sampling frequency differences, downsampling is performed based on the lowest sampling frequency or upsampling is performed based on the highest sampling frequency to ensure that the time dimension of each data type is consistent.

[0043] The synchronized data sets are normalized and data noise filtered to eliminate the dimensional differences of data from different sensors.

[0044] S2: Using sliding window technology and frequency domain analysis algorithm, key indicators such as vibration frequency, temperature fluctuation amplitude and load peak are extracted from the standardized operation feature data set, and the key indicator data are mapped to the equipment operation status space.

[0045] S2.1: Apply sliding window segmentation to the operating feature dataset in time series, define each window as a fixed length, and set the window sliding step to form multiple consecutive time period subsets.

[0046] The data segments within each window are time-aligned again to ensure the synchronization of features in different data dimensions, and the window size and step size are dynamically adjusted according to the device operation standard.

[0047] S2.2: Use fast Fourier transform (FFT) to calculate the main frequency distribution in the window, and extract the first three components with the largest frequency amplitude as vibration characteristics, denoted as F vib ={f1,f2,f3}, where f i is the i-th main frequency component of the vibration signal in the window.

[0048] The temperature fluctuation amplitude ΔT within the statistical window can reflect the thermal change trend.

[0049] The maximum load value P is captured by the data peak recognition algorithm in the sliding window. max and the peak frequency v p , record the correlation with the time series.

[0050] S2.3: Normalize the extracted vibration frequency, temperature fluctuation amplitude and load peak according to the operating standard to generate dimensionless features and construct a multidimensional feature vector [F vib ,ΔT,P max ]Describes the comprehensive operating status of the equipment in the window.

[0051] S2.4: Use the dimensionality reduction technology based on principal component analysis to map the multidimensional feature vector to the two-dimensional plane of the equipment operation state space, define the state point position as [x, y], and store the state point sequence in the feature database in chronological order, and mark the associated index with the original time series.

[0052] Preferably, the multidimensional feature vector [F vib ,ΔT,P max ] is mapped to the two-dimensional coordinates of the running state space as follows:

[0053]

[0054] Where [x,y] T is the two-dimensional coordinate of the running state space, and W is the state mapping weight matrix.

[0055] S3: Based on the extracted key indicator data, multidimensional regression analysis and historical life data are used for training to build an equipment life prediction model.

[0056] The equipment life prediction model dynamically updates weight parameters to adapt to changes in the equipment operating environment and improve prediction accuracy.

[0057] S3.1: Based on the generated two-dimensional plane data of the operating state space, key feature points in the state space are extracted. These feature points include vibration frequency, temperature fluctuation amplitude, load peak value and their position coordinates in the state space.

[0058] Preferably, based on the two-dimensional coordinates [x, y]T Extract distribution characteristic function:

[0059] g(P z )=exp(-λ×||P z -P z,i || 2 )

[0060] Among them, g(P z ) is the distribution influence function of the feature point, P z =[x s ,y s ] T is the real-time running state coordinate, P z,i =[x i ,y i ] T is the historical state coordinate, λ is the distribution function attenuation coefficient, ||P z -P z,i || is the Euclidean distance between the real-time coordinates and the historical coordinates.

[0061] S3.2: Combined with the historical operation data of the equipment, the multidimensional regression analysis method is used to train the extracted two-dimensional plane feature points to build a life prediction model.

[0062] It should be noted that the life prediction model realizes the prediction calculation of equipment life by analyzing the corresponding relationship between the distribution law of characteristic points and equipment life data. The specific formula is as follows:

[0063]

[0064] Where L is the predicted value of the remaining life of the equipment, is a nonlinear mapping function, N is the number of historical feature points, α i is the regression weight of the feature point, β is the regression weight of the historical life data, and H is the historical life data of the equipment.

[0065] S3.3: After the life prediction model training is completed, the feature weight parameters in the life prediction model are dynamically adjusted to adapt to changes in the equipment operating status:

[0066] Δα i =η×||P z -P z,i || ×sign(LL th )

[0067] Among them, Δα i is the weight adjustment amount, η is the adjustment coefficient, sign(x) represents the sign function, and its value is ±1, Lth is the lifespan threshold.

[0068] The adjustment is based on the real-time distribution of feature points in the state space and their deviation from the historical life data.

[0069] S4: Analyze the real-time operation characteristic data of key equipment using the life prediction model, generate maintenance suggestions and mark potential maintenance risks based on the health assessment results and the distribution of real-time characteristic points in the state space.

[0070] Specifically, the real-time operation characteristic data of the key equipment is collected, the data is projected onto the two-dimensional plane of the operation status space in S2.4, and the position coordinates of the latest feature points are calculated; the real-time positions of the feature points are input into the life prediction model, and the remaining life prediction value of the equipment is calculated in combination with the distribution trend of the historical feature points.

[0071] Based on the output of the life prediction model, the current health status of the equipment is evaluated:

[0072]

[0073] Among them, Health is the health index of the device, L y It is the design reference life value of the equipment.

[0074] Furthermore, if the health index Health of the equipment is lower than the health threshold, or the distribution of real-time feature points in the state space deviates from the historical health feature area, maintenance recommendations are generated and potential maintenance risks are marked.

[0075] Specifically, the distribution of real-time feature points in the state space is calculated as follows:

[0076] Calculate the Euclidean distance between the real-time feature points and the historical health area:

[0077]

[0078] Among them, d real is the distance between the real-time feature point and the center of the healthy area, [x he ,y he ] T is the two-dimensional column vector of the center of the healthy area, and is the sum of the squares of the differences of the vector components.

[0079] If the distance d real Exceeding the set radius R of the health zone he , it means that the current state deviates from the healthy area.

[0080] Based on the two trigger conditions, a comprehensive maintenance risk flag is generated, for example:

[0081] If Health <Hth , and d real >R he , then comprehensive maintenance is recommended;

[0082] If Health <H th , then life-related maintenance is recommended;

[0083] If d real >R he , it is recommended to perform a state deviation check.

[0084] Among them, H th is the health threshold.

[0085] S5: Develop an equipment maintenance plan based on the equipment health assessment results and life prediction values, combined with the equipment operation schedule and resource allocation of the park, and dynamically adjust the plan parameters based on the maintenance execution feedback data to reduce equipment downtime.

[0086] According to the equipment health assessment results and remaining life prediction values ​​output by the steps, the maintenance priority of the equipment is divided:

[0087] High-priority equipment: equipment whose predicted remaining life span is significantly lower than the life span threshold, or whose real-time status feature points deviate significantly from the historical healthy area.

[0088] Medium priority device: The device whose predicted lifespan value is close to the lifespan threshold or whose status feature point is near the boundary of the healthy area.

[0089] Low-priority devices: devices with good health status, predicted lifespan values ​​far above the threshold, and feature points stable within the healthy area.

[0090] Combined with the park equipment operation schedule, evaluate the impact of equipment downtime on overall production, and integrate maintenance resources. According to the urgency and complexity of maintenance tasks, reasonably allocate resources in the maintenance team and spare parts library to ensure efficient completion of maintenance tasks.

[0091] During the plan execution process, collect and maintain execution feedback data and dynamically adjust plan parameters, including:

[0092] If the maintenance task exceeds the expected time, analyze the reasons and re-evaluate the maintenance time window of subsequent equipment;

[0093] If resource allocation is insufficient or unreasonable, optimize resource allocation methods to improve the efficiency of subsequent tasks;

[0094] If the health status of the equipment does not improve significantly after repair, the subsequent maintenance strategy will be adjusted and additional diagnostic methods will be introduced.

[0095] Furthermore, this embodiment also provides a park intelligent management system based on multi-source data fusion, including:

[0096] The acquisition module is used to collect vibration data, temperature data and load data of key equipment, synchronize the collected data with the timestamp, and perform normalization and noise filtering according to the key equipment operation standards to form an operation feature data set;

[0097] An extraction module is used to extract key indicators from the standardized operation feature data set using sliding window technology and frequency domain analysis algorithm, and map the key indicator data to the equipment operation state space;

[0098] A module construction module is used to construct an equipment life prediction model based on the extracted key indicator data by using multidimensional regression analysis and historical life data for training;

[0099] An evaluation module, used to analyze the real-time operation characteristic data of key equipment using the life prediction model, generate maintenance suggestions and mark potential maintenance risks based on the health assessment results and the distribution of real-time characteristic points in the state space;

[0100] The maintenance module is used to formulate equipment maintenance plans based on equipment health assessment results and life prediction values, combined with the equipment operation schedule and resource allocation of the park, and dynamically adjust plan parameters based on maintenance execution feedback data to reduce equipment downtime.

[0101] This embodiment also provides a computer device, which is suitable for the case of a campus intelligent management method based on multi-source data fusion, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the campus intelligent management method based on multi-source data fusion as proposed in the above embodiment.

[0102] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0103] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing intelligent management of a campus based on multi-source data fusion as proposed in the above embodiment is implemented.

[0104] In summary, the present invention collects and integrates the vibration data, temperature data and load data of the equipment, uses sliding window technology and frequency domain analysis algorithm to extract key indicators, and trains the life prediction model based on historical life data and multidimensional regression analysis. This comprehensive analysis of multidimensional data can better reflect the health status of the equipment under various operating conditions. At the same time, combined with the operating environment of the equipment and the dynamic update of real-time data, our invention can achieve more accurate equipment life prediction, and generate real-time health assessment and maintenance recommendations based on this, avoiding the limitations of traditional methods that ignore the multidimensional characteristics and dynamic changes of equipment, thereby effectively reducing equipment failure rate and downtime.

[0105] Example 2

[0106] Referring to Table 1, which is the second embodiment of the present invention, this embodiment provides a park intelligent management method based on multi-source data fusion. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0107] In order to verify the effectiveness of the park intelligent management method based on multi-source data fusion, three key equipment (A, B, C) in an industrial park were selected for experiments. Each device is equipped with a vibration sensor, a temperature sensor and a load sensor. The vibration sensor is installed on the bearing, the temperature sensor is arranged in the heat source area, and the load sensor is integrated in the transmission component. The sampling frequencies of the sensors are: 10kHz for vibration signal, 1Hz for temperature signal, and 100Hz for load signal.

[0108] First, the real-time sensor data streams of the three devices are time-stamped and noise-filtered through edge computing devices. For sensor data with significant sampling frequency differences, the highest frequency-based upsampling method is used to unify to 10kHz, and linear interpolation is used to fill the time gaps. The synchronized data sets are normalized to eliminate dimensional differences and form a standardized operational feature data set.

[0109] Then, features were extracted using sliding window technology and frequency domain analysis algorithms. The sliding window was set to 2 seconds and the step size was 1 second. FFT analysis was performed on the vibration signal to extract the top three frequency components. The fluctuation amplitude of temperature data was counted and the peak value of load data was extracted. The extracted eigenvalues ​​were mapped to the equipment operation state space, and principal component analysis (PCA) was used to reduce the multidimensional features to a two-dimensional plane.

[0110] Furthermore, a life prediction model is constructed based on the two-dimensional state space data and the historical life data of the equipment, and the model parameters are dynamically updated. Finally, the life prediction model is used to generate an equipment health assessment report, and maintenance recommendations are given for equipment whose health index is below the threshold. The following is some experimental data:

[0111] Table 1 Experimental data record table

[0112]

[0113] From the above data, we can clearly see that:

[0114] The first three vibration frequency components extracted in the experiment show significant changes in the vibration characteristics during the operation of the equipment. For example, after the operation of equipment A, the vibration frequency 1 increased from 15.3Hz to 18.2Hz, which may be ignored in the existing single sensor data analysis method.

[0115] The data from the temperature sensor shows that the fluctuation amplitude of the heat source area increases after the equipment is put into operation, reflecting the thermal changes of the equipment during long-term operation. Compared with traditional temperature monitoring methods, this method can more comprehensively reflect the temperature characteristics.

[0116] Furthermore, through the two-dimensional state space analysis, the distribution of feature points and their deviation from the historical healthy area can be clearly seen. For example, the feature points of device A after operation are obviously far away from the center of the healthy area.

[0117] The experimental results show that the life prediction model with dynamically updated weights can more accurately predict the remaining life of the equipment. The predicted value of equipment A is reduced from 120 days to 95 days, indicating that maintenance is needed in advance.

[0118] Compared with the existing technology, the present invention overcomes the problems of high dependence on a single data source and low feature extraction efficiency of traditional methods through multi-source data fusion and intelligent analysis. At the same time, through the construction of state space and optimization of life prediction model, it realizes comprehensive control of equipment operation status and accurate health assessment and maintenance recommendations.

[0119] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A park intelligent management method based on multi-source data fusion, characterized by: include: S1: Collect vibration data, temperature data and load data of key equipment, synchronize the collected data with timestamps, and perform normalization and noise filtering according to the key equipment operation standards to form an operation feature data set; S2: Using sliding window technology and frequency domain analysis algorithm, extract key indicators from the standardized operation feature data set, and map the key indicator data to the equipment operation status space; S3: Based on the extracted key indicator data, multidimensional regression analysis and historical life data are used for training to build an equipment life prediction model; S4: Analyze the real-time operation characteristic data of key equipment using the life prediction model, generate maintenance suggestions and mark potential maintenance risks based on the health assessment results and the distribution of real-time characteristic points in the state space; S5: Develop an equipment maintenance plan based on the equipment health assessment results and life prediction values, combined with the equipment operation schedule and resource allocation of the park, and dynamically adjust the plan parameters based on the maintenance execution feedback data to reduce equipment downtime.

2. The park intelligent management method based on multi-source data fusion according to claim 1, characterized in that: The S2 comprises the following steps: Apply sliding window segmentation to the running feature dataset in time series, define each window as a fixed length, and set the window sliding step to form multiple continuous time period subsets; The main frequency distribution in the window is calculated by fast Fourier transform, and the first three components with the largest frequency amplitude are extracted as vibration characteristics, denoted as F vib ={f1,f2,f3}, where f i is the i-th main frequency component of the vibration signal in the window; The temperature fluctuation amplitude ΔT within the statistical window; The maximum load value P is captured by the data peak recognition algorithm in the sliding window. max and peak occurrence frequency; Constructing multidimensional feature vectors [F vib ,ΔT,P max ]Describes the comprehensive operating status of the equipment in the window; The multidimensional feature vector is mapped to a two-dimensional plane of the equipment operation state space by using a dimensionality reduction technique based on principal component analysis. The state point position is defined as [x, y]. The state point sequence is stored in a feature database in chronological order and the associated index with the original time series is marked.

3. The park intelligent management method based on multi-source data fusion as claimed in claim 2, characterized in that: The multidimensional feature vector [F vib ,ΔT,P max ] is mapped to the two-dimensional coordinates of the running state space, and the operation is as follows: Where [x,y] T is the two-dimensional coordinate of the running state space, and W is the state mapping weight matrix.

4. The park intelligent management method based on multi-source data fusion as claimed in claim 3 is characterized by: The S3 comprises the following steps: Based on the generated two-dimensional plane data of the running state space, key feature points in the state space are extracted; Combined with the historical operation data of the equipment, a multidimensional regression analysis method is used to train the extracted two-dimensional plane feature points to build a life prediction model; After the life prediction model training is completed, the feature weight parameters in the life prediction model are dynamically adjusted to adapt to changes in the equipment operating status.

5. The park intelligent management method based on multi-source data fusion according to claim 1, characterized in that: The construction of the life prediction model includes: By analyzing the corresponding relationship between the distribution law of characteristic points and the equipment life data, the prediction calculation of equipment life is realized. The specific formula is as follows: Where L is the predicted value of the remaining life of the equipment, is a nonlinear mapping function, N is the number of historical feature points, α i is the regression weight of the feature point, β is the regression weight of the historical life data, H is the historical life data of the equipment, P z =[x s ,y s ] T It is the real-time running status coordinate; The calculation of the distribution of the real-time feature points in the state space, that is, the calculation of the Euclidean distance d between the real-time feature points and the historical health area real .

6. The park intelligent management method based on multi-source data fusion according to claim 1, characterized in that: The health assessment is calculated as follows: Among them, Health is the health index of the device, L y It is the design reference life value of the equipment.

7. The park intelligent management method based on multi-source data fusion according to claim 6, characterized in that: The generated maintenance recommendations include: If the distance d real Exceeding the set radius R of the health zone he , it means that the current state deviates from the healthy area; Based on the two trigger conditions, a comprehensive maintenance risk marker is generated, including: If Health <H th , and d real >R he , comprehensive maintenance is recommended; If Health <H th , life-related maintenance is recommended; If d real >R he , it is recommended to perform a state deviation check; Among them, H th is the health threshold.

8. A park intelligent management system based on multi-source data fusion, based on the park intelligent management method based on multi-source data fusion according to any one of claims 1 to 7, characterized in that: Also includes: The acquisition module is used to collect vibration data, temperature data and load data of key equipment, synchronize the collected data with the timestamp, and perform normalization and noise filtering according to the key equipment operation standards to form an operation feature data set; An extraction module is used to extract key indicators from the standardized operation feature data set using sliding window technology and frequency domain analysis algorithm, and map the key indicator data to the equipment operation state space; A module construction module is used to construct an equipment life prediction model based on the extracted key indicator data by using multidimensional regression analysis and historical life data for training; An evaluation module, used to analyze the real-time operation characteristic data of key equipment using the life prediction model, generate maintenance suggestions and mark potential maintenance risks based on the health assessment results and the distribution of real-time characteristic points in the state space; The maintenance module is used to formulate equipment maintenance plans based on equipment health assessment results and life prediction values, combined with the equipment operation schedule and resource allocation of the park, and dynamically adjust plan parameters based on maintenance execution feedback data to reduce equipment downtime.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the park intelligent management method based on multi-source data fusion described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the park intelligent management method based on multi-source data fusion described in any one of claims 1 to 7 are implemented.

Images