A hydraulic hoist health state evaluation method based on multi-source information fusion
By using multi-source information fusion and the SVDD method, a four-dimensional health indicator system for hydraulic gate hoists is constructed. Combined with the expert weighting method, real-time and quantitative health assessment of hydraulic gate hoists is achieved, solving the problems of subjectivity and model applicability in existing condition assessments and improving the scientific nature and engineering practicality of the assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- POWERCHINA HUADONG ENG CORP LTD
- Filing Date
- 2026-03-31
- Publication Date
- 2026-07-21
AI Technical Summary
Existing health assessment methods for hydraulic gate hoists rely on single monitoring parameters or human experience, lacking a unified standard for classifying health levels. Traditional physical modeling methods are difficult to apply to complex electromechanical-hydraulic systems, and data-driven methods have shortcomings in the construction of indicator systems and the accuracy of assessments, resulting in highly subjective condition assessment results and difficulty in promoting and applying assessment models.
By integrating multi-source information, monitoring data from hydraulic, mechanical, and electrical systems are collected to construct a four-dimensional health indicator system. A health boundary model is established using the Support Vector Data Description (SVDD) method, and the comprehensive health score is calculated using the expert weighting method, thereby achieving real-time and quantitative assessment of the hydraulic gate hoist.
It enables a comprehensive and objective evaluation of the operating status of hydraulic gate hoists, improves the scientificity and consistency of status assessment, solves the engineering practicality problem of fault identification and early warning, and provides a unified standard for judging health levels.
Smart Images
Figure CN122432960A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of health monitoring technology for hydraulic engineering machinery, specifically relating to a method for evaluating the health status of hydraulic gate hoists based on multi-source information fusion. Background Technology
[0002] Hydraulic gate hoists are key actuators used in water conservancy and hydropower projects to control the opening and closing of gates. Their operational safety directly affects the efficiency of project scheduling, flood control safety, and the overall reliability of the facilities. With the widespread use of hydraulic gate hoists in large and medium-sized hydraulic engineering projects, these devices are prone to problems such as component aging, seal wear, oil contamination, and control malfunctions due to their complex structure, harsh operating environment, and long periods of downtime. Failure to identify equipment degradation trends or potential faults in a timely manner will lead to costly maintenance, significant downtime losses, and even endanger project safety.
[0003] There are three main types of existing health status assessment and fault diagnosis methods:
[0004] 1. Physical model-based approach: This approach derives the performance evolution law by establishing a mathematical model that reflects the physical characteristics and failure mechanism of the equipment. However, for hydraulic gate hoists with complex electromechanical-hydraulic coupling, modeling is difficult and engineering applications are limited.
[0005] 2. Knowledge-based methods: These methods rely on expert experience and heuristic rules, often using expert systems and fuzzy reasoning for diagnosis. However, these methods are heavily influenced by subjective factors and are difficult to adapt to complex and changing working conditions.
[0006] 3. Data-driven approach: This approach utilizes operational monitoring data to conduct health assessments through machine learning and statistical modeling. This method does not require precise physical modeling, is highly adaptable, and has become an important development direction for the health management of hydraulic gate hoists. However, existing research suffers from shortcomings such as insufficient indicator system construction, incomplete data fusion, and inconsistent status level classification.
[0007] Therefore, there is an urgent need for a technical solution that can integrate real-time sensor monitoring data, manual inspection records, and a modeling method based on support vector data description (SVDD) to achieve quantitative assessment of the operating status and health level determination of hydraulic gate hoists, thereby improving the scientific nature and engineering practicality of health management. Summary of the Invention
[0008] To address the aforementioned problems in existing technologies, this invention designs a health status evaluation method for hydraulic gate hoists based on multi-source information fusion. This method addresses the following technical issues in existing hydraulic gate hoist health assessment methods: reliance on single monitoring parameters or human experience leads to highly subjective assessment results and a lack of unified health level classification standards; traditional physical modeling methods are difficult to apply to complex electromechanical-hydraulic systems, making it difficult to widely promote and apply assessment models; and data-driven methods have shortcomings in indicator selection, feature extraction, and assessment accuracy, and cannot comprehensively reflect the operating status of hydraulic gate hoists.
[0009] Therefore, the basic idea of this invention is as follows: First, multi-source monitoring data from the hydraulic, mechanical, and electrical systems are collected and standardized, and a health status index system is established based on statistical principles. On this basis, the Support Vector Data Description (SVDD) method is used to model normally operating samples, forming health boundaries and enabling the identification of abnormal samples. Subsequently, based on different index attributes, three relative degradation calculation models are constructed: "the larger the better," "the smaller the better," and "the optimal middle ground." The index weights are determined using an expert weighting method, ultimately forming a comprehensive health evaluation function. By dividing the comprehensive health status into intervals, the overall state of the hydraulic gate hoist can be classified into three levels: healthy, good, and abnormal, thereby achieving real-time assessment and level determination of the equipment's operating status.
[0010] The technical solution of the present invention is as follows: A method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion, comprising the following steps: A. Collect multi-source monitoring data; Collect multi-source monitoring data, which includes real-time sensor parameters and manual detection information; B. Construct a health indicator system; The collected multi-source monitoring data were preprocessed and standardized, and the 3σ method was used for classification to construct a health indicator system. C. Establish a health boundary model; A health boundary model is established by training a support vector data description method using health operation data. D. Obtain the indicator weights; The relative degradation is calculated based on different indicator attributes, and the indicator weights are obtained by combining expert weighting. E. Calculate the overall health score H; Calculate the overall health score H, and classify the hydraulic gate hoist status as healthy, good, or abnormal based on the threshold range.
[0011] Furthermore, the method for collecting multi-source monitoring data described in step A includes the following steps: A1. Real-time acquisition of sensor parameters; A11. Collect hydraulic system data: Obtain the working pressure and return pressure of the oil circuit through pressure sensors, and measure the flow rate of the main oil circuit and branch circuits using flow meters; monitor the real-time temperature of the oil through temperature sensors, and periodically detect the contamination level and water content through oil samplers.
[0012] A12. Collect mechanical system data: Install displacement sensors on the gate and connecting components of the hydraulic hoist to measure the opening and closing position and stroke deviation of the gate; deploy acceleration sensors on the main stress-bearing parts to obtain vibration amplitude and spectrum information; use force sensors on the guide rail and transmission components to measure frictional resistance; and simultaneously collect the speed curve of the opening and closing process.
[0013] A13. Acquiring Electrical System Data: Motor current is measured in real-time using Hall effect current sensors, voltage is acquired using voltage transformers, temperature is obtained using thermocouples or thermistor sensors, and insulation resistance is periodically checked using insulation monitoring devices. Simultaneously, the response delay and integrity of control signals are recorded by the PLC system to reflect the reliability of the electrical control components.
[0014] A2. Supplement manual testing information; Supplement with manual inspection information, which includes wear on key parts, abnormal noise, oil odor, structural loosening, and oil leakage.
[0015] Furthermore, the method for constructing a health indicator system described in step B includes the following steps: B1. Data preprocessing; B11. Noise Reduction Processing: Wavelet transform or Kalman filtering methods are used to eliminate high-frequency noise during signal acquisition. B12. Outlier Removal: Remove data points that significantly deviate from the normal range using box plots or the 3σ criterion; B13. Imputing missing values: Filling in missing data using time series interpolation or adjacent mean substitution methods; B14. Standardization: Use range standardization or Z-score standardization to map data of different physical quantities to the same dimension range.
[0016] B2. Construct a "four-dimensional, multi-indicator" evaluation system; B21. Determine the four dimensions: The four dimensions include the hydraulic dimension, the mechanical dimension, the electrical dimension, and the supplementary dimension of manual inspection; The hydraulic dimensions include working pressure, oil flow rate, oil temperature, contamination level, hydraulic cylinder displacement, and leakage rate.
[0017] The mechanical dimensions include opening and closing displacement accuracy, opening and closing speed stability, vibration parameters, noise level, wear of key components, and guide rail friction resistance; the vibration parameters include displacement amplitude and root mean square acceleration.
[0018] The electrical dimensions include motor current, voltage stability, motor temperature rise, insulation resistance, control signal status, and operational reliability.
[0019] The supplementary dimensions for manual inspection include wear on key components, abnormal noise, oil odor, and oil leakage.
[0020] B22, Screening multiple indicators; The method for screening multiple indicators includes the following steps: B221. Selection of Indicators: Based on the principles of sensitivity, quantification, and operability, and combined with expert experience and relevant standards, select indicators that significantly characterize the operating status of hydraulic gate hoists. B222, Grading Indicators: Using 3 σ Control charts are used to classify indicators. Let the mean of a certain indicator be... μ Standard deviation is σ Monitoring values x The grading index is then:
[0021] B23. Hierarchical organization of indicators; Organized according to a tree structure of "four dimensions and multiple indicators," a well-defined hierarchical framework for health indicators is formed. The selected indicators include: Hydraulic dimensions: hydraulic pressure, hydraulic flow, oil temperature, oil contamination level, hydraulic cylinder displacement, and hydraulic system leakage rate; Mechanical dimensions: hoist displacement, opening and closing speed, vibration parameters, and structural wear. Electrical dimensions: motor current, motor voltage, motor temperature, control signal status, insulation resistance; Additional manual inspection dimensions: wear of key components, abnormal noise, oil odor, and oil leakage.
[0022] Furthermore, the method for establishing the health boundary model described in step C includes the following steps: C1. Prepare the sample set; Health condition data were filtered through historical operational data and inspection records to construct a training sample set. .in Each sample vector contains data for multiple indicators. N This represents the number of sample vectors.
[0023] C2. Perform feature mapping; Using kernel function The samples are mapped to a high-dimensional feature space, and the kernel function includes a radial basis function (RBF) kernel or a polynomial kernel function. The RBF kernel function is represented as follows:
[0024] in, , The first i The and the first j Two sample vectors Let Variance be the variance.
[0025] C3. Establishing healthy boundaries; The objective function is to find the minimum envelope hypersphere in a high-dimensional space, and the optimization function is:
[0026] The constraints are:
[0027] in, R The radius of the minimum envelope hypersphere, Let C be a slack variable, C be a penalty factor, and a be the center of the minimum envelope hypersphere. For sample vectors The distance to the center of the sphere a. By solving the optimization objective function, the minimum envelope boundary of the healthy sample is obtained.
[0028] C4. Real-time discrimination; After inputting any real-time sample vector z into the health boundary, calculate its distance to the center of the sphere:
[0029] If D(z)≤R, it is considered healthy; if D(z)>R, it is considered abnormal.
[0030] Furthermore, the method for obtaining the index weights described in step D includes the following steps: To quantitatively characterize the degree of deviation of different indicators, a relative degradation function is introduced, which transforms the state of each indicator into a quantified value in the [0,1] interval for easier comprehensive calculation. Based on the different properties of the indicators, three types of degradation functions are designed, as follows: If the indicator is a "the larger the better" type of indicator, then the degradation function is as follows:
[0031] If the indicator is a "the smaller the better" type of indicator, then the degradation function is as follows:
[0032] If the indicator is an intermediate-optimal indicator, then the degradation function is as follows:
[0033] Furthermore, the method for classifying the state of the hydraulic hoist into healthy, good, or abnormal based on a threshold range, as described in step E, includes the following steps: E1, assign weights; An expert weighting method was used, combined with the analysis results, to score each indicator. Key indicators were assigned a score of 10, with the remaining indicators assigned scores of 8, 6, 4, and 2 respectively. After normalization, a weighted set was formed. And satisfy:
[0034] E2. Calculate overall health score H ; Introduce the relative degradation of each indicator. D j The overall health score is defined as:
[0035] in, m This represents the total number of indicators.
[0036] E3. Classification of Health Levels Based on overall health H The overall state of the hydraulic gate hoist is divided into three categories based on the specified range: healthy: H ≤0.3; Good: 0.3 H ≤0.6; abnormal: H >0.6.
[0037] Compared with the prior art, the present invention has achieved the following beneficial effects: 1. This invention integrates multi-source heterogeneous monitoring data from hydraulic, mechanical, and electrical systems, and constructs a hierarchical index system covering four dimensions. This overcomes the limitations of traditional methods that rely on single monitoring parameters or human experience, and achieves a comprehensive and objective evaluation of the operating status of hydraulic gate hoists, thereby improving the scientificity and consistency of the status assessment results.
[0038] 2. This invention adopts the Support Vector Data Description (SVDD) method, which uses only healthy operating samples to establish a minimum envelope hypersphere in a high-dimensional feature space as the health boundary. This solves the engineering problem of scarce fault samples in hydraulic gate hoists and the difficulty in directly applying traditional supervision models, and realizes the identification and early warning of abnormal states under fault-free data conditions.
[0039] 3. This invention designs three types of relative degradation calculation functions for different physical characteristics of indicators: "the larger the better", "the smaller the better", and "the best in the middle". It also combines the expert weighting method to determine the weight of each indicator and finally quantifies the overall status through the comprehensive health function, which solves the shortcomings of existing data-driven methods in terms of indicator selection, feature normalization and comprehensive evaluation accuracy.
[0040] 4. This invention divides the overall health status into three distinct levels: "healthy," "good," and "abnormal." Based on statistical principles, it uses the 3σ method to classify individual indicators, establishing a unified and operable standard for judging health status. This enables real-time and quantitative assessment of the operating status of hydraulic gate hoists, improving the engineering practicality and decision support capabilities of health management. Attached Figure Description
[0041] Figure 1 This is a flowchart of the present invention; Figure 2 This is a technical roadmap of the present invention; Figure 3 This is a framework diagram for constructing the index system of a hydraulic gate hoist according to the present invention; Figure 4 This is a schematic diagram of the 3σ grading determination of a hydraulic gate opener according to the present invention; Figure 5 This is a schematic diagram of an SVDD model according to the present invention; Figure 6 This is a schematic diagram of a kernel function mapping according to the present invention; Figure 7 This is a curve for calculating the intermediate optimal class of degradation according to the present invention; Figure 8 This is a degradation degree calculation curve for a "the larger the better" type of invention; Figure 9 This is a degradation degree calculation curve for a "smaller is better" type according to the present invention; Figure 10 This is a schematic diagram of a weight allocation method according to the present invention; Figure 11 This is a flowchart of the comprehensive health status calculation and level determination of a hydraulic gate hoist according to the present invention. Detailed Implementation
[0042] The technical solution proposed in this application will be described in detail below with reference to the accompanying drawings.
[0043] Figure 1 The flowchart of this invention is as follows: Figure 1 As shown in the flowchart, the health status evaluation method for hydraulic gate hoists includes the following steps: Step 101: Collect multi-source monitoring data, including real-time sensor parameters and manual inspection information. In this embodiment, the operating status information of the hydraulic gate hoist comes from a wide range of sources. To ensure the comprehensiveness and reliability of the data, this step includes the following aspects: Hydraulic system data acquisition: Working and return oil pressures are acquired using pressure sensors; flow meters measure the flow rate in the main and branch oil circuits; temperature sensors monitor the real-time oil temperature; and oil samplers periodically test the contamination level and water content. These parameters reflect the hydraulic system's operational stability, energy transfer efficiency, and oil quality.
[0044] Mechanical system data acquisition: Displacement sensors are installed on the gate and connecting components of the hoist to measure the opening and closing position and stroke deviation of the gate; acceleration sensors are deployed at the main stress-bearing parts to acquire vibration amplitude and spectrum information; force sensors are used on the guide rails and transmission components to measure frictional resistance; and velocity curves during the opening and closing process are simultaneously acquired. These indicators reflect the precision retention, structural wear, and motion stability of the mechanical system.
[0045] Electrical system data acquisition: Motor current is measured in real time using a Hall current sensor, voltage is acquired using a voltage transformer, temperature is obtained using a thermocouple or thermistor sensor, and insulation resistance is periodically monitored by an insulation monitoring device. Simultaneously, the PLC system records the response delay and integrity of control signals to reflect the reliability of the electrical control components.
[0046] Manual inspection data supplement: Manual inspection information supplements fault signs in specific scenarios, such as noise, odor, structural abnormalities, oil performance degradation curves, cumulative wear, etc. Step 102: Preprocess and standardize the collected data, use the 3σ method for classification, and construct a health indicator system. After data collection is completed, preprocessing is required: Denoising: Wavelet transform or Kalman filtering methods are used to eliminate high-frequency noise during signal acquisition. Outlier removal: Data points that significantly deviate from the normal range are removed using box plots or the 3σ criterion; Missing value imputation: Filling in missing data using time series interpolation or adjacent mean substitution methods; Standardization: Range standardization or Z-score standardization is used to map data of different physical quantities to the same dimension range, ensuring the computational stability of subsequent models.
[0047] Through the above steps, a high-quality, multi-dimensional, and continuously usable hydraulic gate hoist operation database is established, providing reliable input for subsequent indicator system construction and model training.
[0048] like Figure 3 As shown, this embodiment constructs a "four-dimensional—multi-indicator" evaluation system: Hydraulic dimensions include indicators such as hydraulic pressure, hydraulic flow rate, oil temperature, oil contamination level, hydraulic cylinder displacement, and hydraulic system leakage rate. These indicators reflect the efficiency of hydraulic energy transmission, system stability, and oil quality, and are key indicators of the health of the hydraulic system.
[0049] Mechanical dimension: This includes opening and closing displacement, opening and closing speed stability, vibration parameters (displacement amplitude, root mean square acceleration), and structural wear. This dimension primarily reflects the reliability and motion performance of the mechanical structure.
[0050] Electrical dimension: This includes motor current, motor voltage, motor temperature rise, insulation resistance, and control signal status. This dimension measures the stability and safety of the electrical drive and control components.
[0051] Additional dimensions for manual inspection: wear on key components, abnormal noise, unusual oil odor, and oil leakage. This dimension supplements the three dimensions mentioned above with crucial information.
[0052] To ensure the scientific validity of the indicator system, the following method is adopted in this embodiment: Indicator selection: Based on the principles of sensitivity, quantifiability, and operability, and combined with expert experience and relevant standards, indicators that can significantly characterize the system's operating status are selected. Indicator classification: Classification is performed using the 3σ control chart method. For example... Figure 4 As shown, let the mean of a certain indicator be μ, the standard deviation be σ, and the monitoring value be μ. x Then it is divided into:
[0053] This method is based on statistics and can dynamically reflect data fluctuations, avoiding reliance on a single empirical threshold.
[0054] Hierarchical organization of indicators: The above indicators are organized according to a tree structure of "dimension-sub-indicator" to form a clear hierarchical indicator system framework, which ensures comprehensiveness and facilitates subsequent modeling and calculation.
[0055] Step 103: Using health operation data, train the model using the SVDD method to establish a health boundary model. After establishing the indicator system, it is necessary to further model the overall health boundary of the hydraulic gate hoist. Since fault samples are extremely rare or even nonexistent during the operation of the hydraulic gate hoist, traditional supervised classification methods are difficult to apply. Therefore, this embodiment uses the Support Vector Data Description (SVDD) method, constructing an "envelope boundary" using only healthy samples, and using this to determine the health status of new samples.
[0056] 1. Sample set preparation: Health condition data were filtered through historical operational data and inspection records to construct a training sample set. .in For each sample vector, there are multiple indicator data; N is the number of sample vectors. These samples have been standardized to eliminate differences in units and ensure comparability between different indicators.
[0057] 2. Feature mapping: Using kernel function This maps samples to a high-dimensional feature space. Commonly used kernel functions include radial basis function (RBF) kernels and polynomial kernels. This embodiment uses the RBF kernel function:
[0058] in, , Given two sample vectors, The variance is denoted as σ; this kernel function can capture nonlinear characteristic relationships and is suitable for complex electromechanical-hydraulic coupled systems such as hydraulic gate hoists.
[0059] 3. Establishing Health Boundaries: The objective function is to find the minimum envelope hypersphere in a high-dimensional space, and the optimization function is:
[0060] The constraints are:
[0061] in, R For radius, Let C be a slack variable, C be a penalty factor, and a be the center of the sphere. For sample vectors The distance to the center of the sphere a. By solving this optimization objective function, the minimum envelope boundary of the healthy samples is obtained.
[0062] Real-time identification: like Figure 5 and Figure 6 As shown, after any real-time sample vector z is input into the model, its distance to the center of the sphere is calculated:
[0063] If D(z) ≤ R, the device is considered healthy; if D(z) > R, it is considered abnormal. This method can detect the degree of deviation of the device's status in real time.
[0064] By using SVDD modeling, this embodiment solves the problem of "scarcity of fault samples" and realizes adaptive boundary modeling based on health data, laying the foundation for subsequent quantitative assessment.
[0065] Step 104: Calculate the relative degradation degree based on different indicator attributes, and obtain the indicator weights by combining expert weighting. While SVDD can identify anomalies, it cannot quantitatively characterize the degree of deviation of different indicators. Therefore, this embodiment introduces a relative degradation function to convert the state of each indicator into a quantitative value in the [0,1] interval, which facilitates comprehensive calculation.
[0066] Degradation function design: Based on the different properties of the indicators, three types of degradation functions are designed: 1. Indicators that prioritize larger values:
[0067] When the index value is close to the standard value, the degree of degradation is close to 0; as the value decreases, the degree of degradation increases monotonically.
[0068] 2. Indicators where smaller is better:
[0069] When the index value is close to the minimum, the degradation degree is close to 0; as the value increases, the degradation degree increases.
[0070] 3. Intermediate-optimal index:
[0071] When the index value equals the standard value, the degree of degradation is 0; regardless of how much it deviates from the standard value, the degree of degradation increases, showing a "U-shaped curve".
[0072] Degradation curve representation: like Figures 7 to 9 As shown, the three types of functions correspond to decreasing curves, increasing curves, and U-shaped curves, respectively. These curves provide a clear visual representation of the health deviation patterns of different types of indicators.
[0073] Step 105: Calculate the overall health score H, and classify the gate operation status into healthy, good, or abnormal based on the threshold range. Since different indicators have varying degrees of importance to the overall health status of the hydraulic gate hoist, failing to weight them may obscure key indicators. Therefore, this embodiment introduces a weighting method and calculates the overall health score.
[0074] Weighting: An expert weighting method was used to score each indicator. Key indicators were assigned a score of 10, with the remaining indicators assigned scores of 8, 6, 4, and 2 respectively. After normalization, a weighted set was formed. And satisfy:
[0075] Overall health score calculation: Introduce the relative degradation of each indicator. Overall health H Defined as:
[0076] Where m represents the total number of indicators.
[0077] Health level classification: like Figure 10 and Figure 11 As shown, based on overall health H The overall state of the hydraulic gate hoist is divided into three categories based on the specified range: healthy: H ≤0.3 Good: 0.3 H ≤0.6 abnormal: H >0.6 Through this step, the present invention achieves quantitative integration from single indicators to overall health, ensuring that the diagnostic results reflect both individual indicator abnormalities and maintain the scientific nature of the overall assessment.
[0078] This invention is not limited to this embodiment. Any equivalent concept or modification within the technical scope disclosed in this invention shall be included within the protection scope of this invention.
Claims
1. A method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion, characterized in that: Includes the following steps: A. Collect multi-source monitoring data; Collect multi-source monitoring data, which includes real-time sensor parameters and manual detection information; B. Construct a health indicator system; The collected multi-source monitoring data were preprocessed and standardized, and the 3σ method was used for classification to construct a health indicator system. C. Establish a health boundary model; A health boundary model is established by training a support vector data description method using health operation data. D. Obtain the indicator weights; The relative degradation is calculated based on different indicator attributes, and the indicator weights are obtained by combining expert weighting. E. Calculate the overall health score H; Calculate the overall health score H, and classify the hydraulic gate hoist status as healthy, good, or abnormal based on the threshold range.
2. The method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion as described in claim 1, characterized in that: The method for collecting multi-source monitoring data described in step A includes the following steps: A1. Real-time acquisition of sensor parameters; A11. Collect hydraulic system data: Obtain the working pressure and return pressure of the oil circuit through pressure sensors, and measure the flow rate of the main oil circuit and branch circuits using flow meters; monitor the real-time temperature of the oil through temperature sensors, and periodically detect the contamination level and water content through an oil sampler. A12. Collect mechanical system data: Install displacement sensors on the gate and connecting components of the hydraulic hoist to measure the opening and closing position and stroke deviation of the gate; deploy acceleration sensors on the main stress-bearing parts to obtain vibration amplitude and spectrum information; use force sensors on the guide rail and transmission components to measure frictional resistance; and simultaneously collect the speed curve of the opening and closing process. A13. Acquiring electrical system data: Real-time measurement of motor current using Hall current sensors, voltage acquisition using voltage transformers, temperature acquisition using thermocouples or thermistor sensors, and periodic detection of insulation resistance using insulation monitoring devices; Simultaneously, recording of control signal response delay and integrity using the PLC system to reflect the reliability of the electrical control links; A2. Supplement manual testing information; Supplement with manual inspection information, which includes wear on key parts, abnormal noise, oil odor, structural loosening, and oil leakage.
3. The method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion according to claim 1, characterized in that: The method for constructing a health indicator system described in step B includes the following steps: B1. Data preprocessing; B11. Noise Reduction Processing: Wavelet transform or Kalman filtering methods are used to eliminate high-frequency noise during signal acquisition. B12. Outlier Removal: Remove data points that significantly deviate from the normal range using box plots or the 3σ criterion; B13. Imputing missing values: Filling in missing data using time series interpolation or adjacent mean substitution methods; B14. Standardization: Use range standardization or Z-score standardization to map data of different physical quantities to the same dimension range. B2. Construct a "four-dimensional, multi-indicator" evaluation system; B21. Determine the four dimensions: The four dimensions include the hydraulic dimension, the mechanical dimension, the electrical dimension, and the supplementary dimension of manual inspection; The hydraulic dimensions include working pressure, oil flow rate, oil temperature, contamination level, hydraulic cylinder displacement, and leakage rate. The mechanical dimensions include opening and closing displacement accuracy, opening and closing speed stability, vibration parameters, noise level, wear of key components, and guide rail friction resistance; the vibration parameters include displacement amplitude and root mean square acceleration. The electrical dimensions include motor current, voltage stability, motor temperature rise, insulation resistance, control signal status, and operational reliability. The supplementary dimensions for manual inspection include wear on key components, abnormal noise, unusual oil odor, and oil leakage. B22, Screening multiple indicators; The method for screening multiple indicators includes the following steps: B221. Selection of Indicators: Based on the principles of sensitivity, quantification, and operability, and combined with expert experience and relevant standards, select indicators that significantly characterize the operating status of hydraulic gate hoists. B222, Grading Indicators: Using 3 σ Control charts are used to classify indicators; let the mean of a certain indicator be... μ Standard deviation is σ Monitoring values x The grading index is then: B23. Hierarchical organization of indicators; Organized according to a tree structure of "four dimensions and multiple indicators," a well-defined hierarchical framework for health indicators is formed; the selected indicators include: Hydraulic dimensions: hydraulic pressure, hydraulic flow, oil temperature, oil contamination level, hydraulic cylinder displacement, and hydraulic system leakage rate; Mechanical dimensions: hoist displacement, opening and closing speed, vibration parameters, and structural wear. Electrical dimensions: motor current, motor voltage, motor temperature, control signal status, insulation resistance; Additional manual inspection dimensions: wear of key parts, abnormal noise, oil odor, and oil leakage.
4. The method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion according to claim 1, characterized in that: The method for establishing a health boundary model as described in step C includes the following steps: C1. Prepare the sample set; Health condition data were filtered through historical operational data and inspection records to construct a training sample set. ;in Each sample vector contains data for multiple indicators. N The number of sample vectors; C2. Perform feature mapping; Using kernel function The samples are mapped to a high-dimensional feature space. The kernel function includes a radial basis function (RBF) kernel function or a polynomial kernel function. The RBF kernel function is represented as follows: in, , The first i The and the first j Two sample vectors For variance; C3. Establishing healthy boundaries; The objective function is to find the minimum envelope hypersphere in a high-dimensional space, and the optimization function is: The constraints are: in, R The radius of the minimum envelope hypersphere, Let C be the slack variable, C be the penalty factor, and a be the center of the minimum envelope hypersphere. For sample vectors The distance to the center of the sphere a; by solving the optimization objective function, the minimum envelope boundary of the healthy sample is obtained; C4. Real-time discrimination; After inputting any real-time sample vector z into the health boundary, calculate its distance to the center of the sphere: If D(z)≤R, it is considered healthy; if D(z)>R, it is considered abnormal.
5. The method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion according to claim 1, characterized in that: The method for obtaining indicator weights as described in step D includes the following steps: To quantitatively characterize the degree of deviation of different indicators, a relative degradation function is introduced, which transforms the state of each indicator into a quantified value in the [0,1] interval for easier comprehensive calculation. Based on the different properties of the indicators, three types of degradation functions are designed, as follows: If the indicator is a "the larger the better" type of indicator, then the degradation function is as follows: If the indicator is a "the smaller the better" type of indicator, then the degradation function is as follows: If the indicator is an intermediate-optimal indicator, then the degradation function is as follows: 。 6. The method for evaluating the health status of a hydraulic gate hoist based on multi-source information fusion according to claim 1, characterized in that: Step E, which describes the method for classifying the state of a hydraulic gate hoist into healthy, good, or abnormal based on a threshold range, includes the following steps: E1, assign weights; An expert weighting method was used, combined with the analysis results, to score each indicator; the key indicator was assigned a score of 10, and the remaining indicators were assigned scores of 8, 6, 4, and 2 respectively; after normalization, a weight set was formed. And satisfy: E2. Calculate overall health score H ; Introduce the relative degradation of each indicator. D j The overall health score is defined as: in, m The total number of indicators; E3. Classification of Health Levels Based on overall health H The overall state of the hydraulic gate hoist is divided into three categories based on the specified range: healthy: H ≤0.3; Good: 0.3 H ≤0.6; abnormal: H >0.6.