Equipment health state fusion evaluation method based on multi-sensor data

Abstract

The invention belongs to the technical field of intelligent operation and maintenance, and discloses an equipment health state fusion evaluation method based on multi-sensor data. Comprising the following steps: acquiring fan equipment working condition data in a monitoring time period; classifying the working condition data of the fan equipment according to a logic causal relationship among the working condition data of the fan equipment to obtain working condition data of a power supply side and working condition data of a machine side; according to the power supply side working condition data and the mechanical side working condition data, the health state of the fan equipment is evaluated, and a first evaluation state and a second evaluation state of the fan equipment are obtained; when the first evaluation state and the second evaluation state of the fan equipment are both healthy states, the fan equipment is in a healthy state, and when the first evaluation state of the fan equipment is an unhealthy state or the second evaluation state of the fan equipment is an unhealthy state, the fan equipment is not in a healthy state; and the accuracy of health assessment of the fan equipment and the timeliness of maintenance of the fan equipment are further improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance technology, and more specifically, to a method for evaluating the health status of equipment based on multi-sensor data fusion. Background Technology

[0002] Patent application CN119226991A discloses a health status assessment system and method for a main helium blower based on state estimation, relating to the field of nuclear power plant equipment condition monitoring and health management technology. The system includes a data acquisition unit, a feature extraction unit, a state estimation unit, and a health assessment unit. It employs strategies such as multi-sensor data fusion, mechanistic data hybrid modeling, and active incremental learning to construct an end-to-end health status assessment model; and introduces blockchain technology to achieve secure data sharing and trusted computing. This system can comprehensively utilize multi-source heterogeneous data to automatically, in real-time, and accurately assess the health status of the main helium blower, providing a reliable basis for intelligent operation and maintenance decisions. This improves the safety and economy of nuclear power plants and promotes the intelligent development of nuclear power equipment.

[0003] However, in assessing the health status of wind turbine equipment, traditional methods typically set thresholds to determine whether relevant parameters are normal. However, significant fluctuations in the current, voltage, or temperature of the wind turbine equipment can indicate potential malfunctions and unhealthy conditions. Furthermore, there is a correlation between the current, voltage, and temperature of the wind turbine equipment; analyzing only one parameter may lead to inaccurate health status assessments, increasing plant operating costs and reducing profits. Additionally, dust exists inside the wind turbine equipment, and as the equipment operates for an extended period, this dust can unevenly accumulate on the impeller. Gradually disrupting the dynamic balance of the fan equipment increases the bearing amplitude. Traditional methods typically measure the bearing amplitude and stop the fan to clean the dust when it reaches the maximum allowable value. However, by the time the bearing amplitude reaches the maximum allowable value, the fan is already in a dangerous state. Failure to stop the fan in advance could damage it and cause a safety accident. Traditional methods cannot provide real-time information on the time required for the bearing amplitude to reach the maximum allowable value, thus preventing timely shutdown for dust cleaning. This increases the risk of fan malfunction and damage, and raises the probability of safety accidents.

[0004] In view of this, the present invention proposes a device health status fusion assessment method based on multi-sensor data to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a device health status fusion assessment method based on multi-sensor data, comprising: Step S1: Obtain wind turbine equipment operating data within the monitoring period using heterogeneous sensors; Step S2: Classify the wind turbine equipment operating data according to the logical causal relationship between the operating data to obtain power supply side operating data and mechanical side operating data; Step S3: Construct a combined three-Y-axis line graph of current-voltage-bearing temperature based on the power supply side operating data. Obtain the current factor, voltage factor, and bearing temperature factor through the combined three-Y-axis line graph of current-voltage-bearing temperature. Calculate the weighted average of the current factor, voltage factor, and bearing temperature factor to obtain the health assessment factor of the wind turbine equipment. Evaluate the health status of the wind turbine equipment based on the health assessment factor of the wind turbine equipment and the combined three-Y-axis line graph of current-voltage-bearing temperature to obtain the first assessment status of the wind turbine equipment. Step S4: Based on the mechanical side operating condition data, determine whether the bearing amplitude of the fan equipment is normal and whether the sound signal generated by the fan equipment is normal. Based on the judgment results, obtain the second evaluation status of the fan equipment. Based on the first and second evaluation statuses of the fan equipment, determine whether the fan equipment is in a healthy state. When the bearing amplitude of the fan equipment is normal, obtain the time required for the bearing amplitude to change from normal to abnormal, and perform fan equipment maintenance in advance based on the obtained time.

[0006] Furthermore, the operating data of the wind turbine equipment includes the current and voltage of the internal circuits of the wind turbine equipment, the bearing amplitude and bearing temperature of the wind turbine equipment, and the sound signals generated during the operation of the wind turbine equipment; The power supply side operating data includes the current and voltage of the internal circuits of the wind turbine equipment, as well as the bearing temperature of the wind turbine equipment. The mechanical side operating data includes the bearing amplitude of the wind turbine equipment and the sound signals generated during the operation of the wind turbine equipment.

[0007] Furthermore, the method for constructing a combined current-voltage-bearing temperature three-Y-axis line graph based on power supply side operating condition data includes: The monitoring period is evenly divided into u monitoring time points, and the current, voltage of the internal circuits of the wind turbine equipment and the bearing temperature of the wind turbine equipment are obtained at each monitoring time point. Establish a blank three-Y coordinate system. Set the horizontal axis of the blank three-Y coordinate system to time, set the first Y axis of the blank three-Y coordinate system to current, set the second Y axis of the blank three-Y coordinate system to voltage, and set the third Y axis of the blank three-Y coordinate system to temperature. The current, voltage, and bearing temperature of the internal circuits of the wind turbine equipment at each monitoring time point are filled into a blank three-Y-axis coordinate system. The current, voltage, and bearing temperature markers at each monitoring time point are obtained respectively. Straight lines are drawn in chronological order to connect the current markers, voltage markers, and bearing temperature markers in chronological order, thus obtaining a combined three-Y-axis polygonal graph of current-voltage-bearing temperature.

[0008] Furthermore, the method for assessing the health status of the wind turbine equipment includes: The first assessment status includes a healthy state and an unhealthy state; Set current threshold, voltage threshold and bearing temperature threshold. When the current in the internal circuit of the fan equipment corresponding to the current mark point is greater than or equal to the current threshold, or the voltage in the internal circuit of the fan equipment corresponding to the voltage mark point is greater than or equal to the voltage threshold, or the bearing temperature of the fan equipment corresponding to the bearing temperature mark point is greater than or equal to the bearing temperature threshold, the first evaluation state of the fan equipment is unhealthy. When the current in the internal circuit of the fan equipment corresponding to the current mark point is less than the current threshold, the voltage in the internal circuit of the fan equipment corresponding to the voltage mark point is less than the voltage threshold, and the bearing temperature of the fan equipment corresponding to the bearing temperature mark point is less than the bearing temperature threshold, the health assessment factor of the fan equipment is obtained. Set a threshold for the health assessment factor of the wind turbine equipment. If all the health assessment factors of the wind turbine equipment are less than the threshold, the first assessment state of the wind turbine equipment is healthy. If there is a health assessment factor of the wind turbine equipment that is greater than or equal to the threshold, the first assessment state of the wind turbine equipment is unhealthy.

[0009] Furthermore, the method for obtaining the health assessment factors of the wind turbine equipment includes: Starting from the second monitoring time point, draw a perpendicular line to the horizontal axis through the previous current marker, the previous voltage marker, and the previous bearing temperature marker, and record it as the first perpendicular line; Draw a line parallel to the horizontal axis through the current marking point and record it as the current parallel line; draw a line parallel to the horizontal axis through the voltage marking point and record it as the voltage parallel line; draw a line parallel to the horizontal axis through the bearing temperature marking point and record it as the bearing temperature parallel line; record the intersection of the current parallel line and the first perpendicular line as the current intersection point; record the intersection of the voltage parallel line and the first perpendicular line as the voltage intersection point; record the intersection of the bearing temperature parallel line and the first perpendicular line as the bearing temperature intersection point. Obtain the area of ​​the figure bounded by the straight line between the current marker point and the current intersection point, the straight line between the current marker point and the previous current marker point, and the straight line between the previous current marker point and the current intersection point, and denot it as the current factor. Obtain the area of ​​the figure bounded by the straight line between the voltage marker point and the voltage intersection point, the straight line between the voltage marker point and the previous voltage marker point, and the straight line between the previous voltage marker point and the voltage intersection point, and denot it as the voltage factor; Obtain the area of ​​the figure enclosed by the straight line between the bearing temperature mark point and the bearing temperature intersection point, the straight line between the bearing temperature mark point and the previous bearing temperature mark point, and the straight line between the previous bearing temperature mark point and the bearing temperature intersection point, and record it as the bearing temperature factor. The weighted average of the current factor, voltage factor, and bearing temperature factor is calculated, and the calculated weighted average is used as the health assessment factor of the wind turbine equipment at the monitoring time point until the end of the last monitoring time point.

[0010] Furthermore, the method for determining whether the bearing amplitude of the fan equipment is normal and whether the sound signal generated by the fan equipment is normal based on mechanical side operating condition data includes: Set a bearing amplitude threshold. When the bearing amplitude of the fan equipment is greater than or equal to the bearing amplitude threshold, the bearing amplitude of the fan equipment is abnormal. When the bearing amplitude of the fan equipment is less than the bearing amplitude threshold, the bearing amplitude of the fan equipment is normal. The sound signal generated during the operation of the wind turbine is converted to the frequency domain by using Fast Fourier Transform, and the spectrum diagram of the sound signal generated during the operation of the wind turbine is obtained. The spectrum of the sound signal generated during the operation of the wind turbine is input into the wind turbine fault prediction model based on the convolutional neural network to obtain the output result of the wind turbine fault prediction model. The input of the wind turbine fault prediction model is the spectrum, and the output is whether the sound signal generated by the wind turbine is normal or abnormal.

[0011] Furthermore, the method for obtaining the second evaluation state of the wind turbine equipment based on the judgment result includes: The second assessment status includes a healthy state and an unhealthy state; If the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is normal and the bearing amplitude of the wind turbine equipment is normal, then the second evaluation state of the wind turbine equipment is a healthy state. If the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is abnormal or the bearing amplitude of the wind turbine equipment is abnormal, then the second evaluation state of the wind turbine equipment is an unhealthy state.

[0012] Furthermore, the method for determining whether the wind turbine equipment is in a healthy state based on the first and second assessment states of the wind turbine equipment includes: The wind turbine is in a healthy state when both its first and second assessment states are healthy. The wind turbine is not in a healthy state when either its first or second assessment state is unhealthy.

[0013] Furthermore, the method for obtaining the time required for the bearing amplitude to change from normal to abnormal when the bearing amplitude of the wind turbine equipment is normal, and for performing wind turbine equipment maintenance in advance based on the obtained time, includes: Under different dust concentration conditions, the dust concentration refers to the dust concentration in the air inside the fan equipment. The fan equipment is simulated and the mass of dust adsorbed by the impeller and the bearing amplitude of the fan equipment are recorded every p minutes. The mass of dust adsorbed by the impeller of the fan equipment is fitted with time to obtain the fitting function corresponding to the mass of dust adsorbed by the impeller of the fan equipment with time, and is denoted as the mass-time fitting function. The mass of dust adsorbed by the impeller and the bearing amplitude of the fan equipment are fitted to obtain the fitting function corresponding to the mass of dust adsorbed by the impeller and the bearing amplitude of the fan equipment, and are denoted as the mass-amplitude fitting function. The time required for the bearing amplitude of the fan equipment to reach the bearing amplitude threshold is obtained based on the mass-time fitting function and the mass-amplitude fitting function. The fan equipment is stopped q minutes before the bearing amplitude reaches the bearing amplitude threshold, and the dust adsorbed by the impeller of the fan equipment is cleaned.

[0014] Furthermore, the method for obtaining the time required for the bearing amplitude of the wind turbine to reach the bearing amplitude threshold based on the mass-time fitting function and the mass-amplitude fitting function includes: The dust concentration in the air inside the fan equipment is obtained and recorded as the real-time dust concentration. The bearing amplitude threshold is input into the mass-amplitude fitting function corresponding to the real-time dust concentration to obtain the dust mass adsorbed by the impeller of the fan equipment corresponding to the bearing amplitude threshold, which is recorded as the maximum dust mass. The maximum dust mass is input into the mass-time fitting function corresponding to the real-time dust concentration to obtain the time corresponding to the maximum dust mass. The bearing amplitude of the fan equipment is input into the mass-amplitude fitting function corresponding to the real-time dust concentration to obtain the dust mass adsorbed by the impeller of the fan equipment corresponding to the bearing amplitude of the fan equipment, and recorded as the real-time dust mass. The real-time dust mass is then input into the mass-time fitting function corresponding to the real-time dust concentration to obtain the time corresponding to the real-time dust mass. Subtract the time corresponding to the real-time dust mass from the time corresponding to the maximum dust mass to obtain the time required for the bearing amplitude of the fan equipment to reach the bearing amplitude threshold.

[0015] The technical effects and advantages of the device health status fusion assessment method based on multi-sensor data of the present invention are as follows: 1. Based on the logical causal relationship between the operating data of wind turbine equipment, the operating data of wind turbine equipment is classified to obtain the operating data of the power supply side and the mechanical side. There are correlations between the operating data of wind turbine equipment. By jointly analyzing the operating data of wind turbine equipment with correlations, the accuracy of the health status assessment of wind turbine equipment is greatly increased, thereby reducing the operating cost of the factory and increasing the factory's revenue. 2. The health status of the fan equipment is assessed based on the health evaluation factors and the three Y-axis broken line of current-voltage-bearing temperature. The health status of the fan equipment is assessed more comprehensively by combining the changes in current, voltage and temperature of the fan equipment, thereby improving the accuracy of the health status assessment of the fan equipment, reducing the operating costs of the factory and increasing the factory's profits. 3. When the bearing amplitude of the fan equipment is normal, obtain the time required for the bearing amplitude to change from normal to abnormal, and perform fan equipment maintenance in advance based on the obtained time. By monitoring the time required for the bearing amplitude to change from normal to abnormal in real time, the fan equipment can be shut down and cleaned of dust before it is in a dangerous state, thereby reducing the risk of damage caused by fan equipment failure and reducing the probability of safety accidents. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of a device health status fusion assessment method based on multi-sensor data according to the present invention; Figure 2 This is a schematic diagram of a device health status fusion assessment system based on multi-sensor data according to the present invention; Figure 3 This is a flowchart of the method for assessing the health status of wind turbine equipment according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 Please see Figure 1 and Figure 3 As shown, this invention discloses a device health status fusion assessment method based on multi-sensor data, comprising: Step S1: Obtain wind turbine equipment operating data within the monitoring period using heterogeneous sensors; Step S2: Classify the wind turbine equipment operating data according to the logical causal relationship between the operating data to obtain power supply side operating data and mechanical side operating data; Step S3: Construct a combined three-Y-axis line graph of current-voltage-bearing temperature based on the power supply side operating data. Obtain the current factor, voltage factor, and bearing temperature factor through the combined three-Y-axis line graph of current-voltage-bearing temperature. Calculate the weighted average of the current factor, voltage factor, and bearing temperature factor to obtain the health assessment factor of the wind turbine equipment. Evaluate the health status of the wind turbine equipment based on the health assessment factor of the wind turbine equipment and the combined three-Y-axis line graph of current-voltage-bearing temperature to obtain the first assessment status of the wind turbine equipment. Step S4: Based on the mechanical side operating condition data, determine whether the bearing amplitude of the fan equipment is normal and whether the sound signal generated by the fan equipment is normal. Based on the judgment results, obtain the second evaluation status of the fan equipment. Based on the first and second evaluation statuses of the fan equipment, determine whether the fan equipment is in a healthy state. When the bearing amplitude of the fan equipment is normal, obtain the time required for the bearing amplitude to change from normal to abnormal, and perform fan equipment maintenance in advance based on the obtained time.

[0019] The process of acquiring wind turbine equipment operating condition data within a monitoring period using heterogeneous sensors includes: The operating data of the wind turbine equipment includes the current and voltage of the internal circuits of the wind turbine equipment, the bearing amplitude and bearing temperature of the wind turbine equipment, and the sound signals generated during the operation of the wind turbine equipment; Heterogeneous sensors include current sensors, voltage sensors, vibration sensors, temperature sensors, and acoustic sensors; Current sensors are used to acquire the current in the internal circuits of the wind turbine equipment; voltage sensors are used to acquire the voltage in the internal circuits of the wind turbine equipment; vibration sensors are used to acquire the bearing amplitude of the wind turbine equipment; temperature sensors are used to acquire the bearing temperature of the wind turbine equipment; and acoustic sensors are used to acquire the sound signals generated during the operation of the wind turbine equipment. Install current sensors, voltage sensors, vibration sensors, temperature sensors, and acoustic sensors at the appropriate locations on the fan equipment; After installation, current sensors, voltage sensors, vibration sensors, temperature sensors, and acoustic sensors are used to acquire the current and voltage of the internal circuits of the wind turbine equipment, the bearing amplitude and bearing temperature of the wind turbine equipment, and the sound signals generated during the operation of the wind turbine equipment during the monitoring period.

[0020] The process of classifying wind turbine equipment operating data according to the logical causal relationships between the data, and obtaining power supply side operating data and mechanical side operating data, includes: If the voltage of the internal circuit of the wind turbine equipment decreases while the load of the wind turbine equipment remains unchanged, the current of the internal circuit of the wind turbine equipment will increase to maintain the output power. The increase in the current of the internal circuit of the wind turbine equipment will also lead to increased heating of the internal coil of the wind turbine equipment. The heat is conducted to the bearings of the wind turbine equipment, causing the bearing temperature of the wind turbine equipment to increase. Therefore, there is a logical causal relationship between the current, voltage of the internal circuit of the wind turbine equipment and the bearing temperature of the wind turbine equipment. The current, voltage of the internal circuit of the wind turbine equipment and the bearing temperature of the wind turbine equipment are analyzed as a whole, and the current, voltage of the internal circuit of the wind turbine equipment and the bearing temperature of the wind turbine equipment are recorded as the power supply side operating condition data. If the bearing amplitude of the fan equipment increases, acoustic abnormalities such as metal knocking, resonance, and loosening are likely to occur. Therefore, there is a logical causal relationship between the bearing amplitude of the fan equipment and the sound signals generated during the operation of the fan equipment. The bearing amplitude of the fan equipment and the sound signals generated during the operation of the fan equipment are analyzed as a whole, and the bearing amplitude of the fan equipment and the sound signals generated during the operation of the fan equipment are recorded as mechanical side operating condition data.

[0021] The process of constructing a combined three-Y-axis line graph of current, voltage, and bearing temperature based on power supply side operating data includes: The first assessment status includes healthy status and unhealthy status; The monitoring period is evenly divided into u monitoring time points. u can be set by experimental data analysis or experience, and the current, voltage and bearing temperature of the internal circuit of the fan equipment and the bearing temperature of the fan equipment are obtained at each monitoring time point. Establish a blank three-Y-axis coordinate system. Set the x-axis of the blank three-Y-axis coordinate system to time, the first Y-axis to current, the second Y-axis to voltage, and the third Y-axis to temperature. Highlight the first, second, and third Y-axis in red, yellow, and green, respectively. Fill the blank three-Y-axis coordinate system with the current, voltage, and bearing temperature of the wind turbine equipment's internal circuits at each monitoring time point, and obtain the current index at each monitoring time point. The current, voltage, and bearing temperature markings are plotted in chronological order. Straight lines are drawn to connect the current markings sequentially, and the lines connecting the current markings and the current markings are highlighted in red. Straight lines are drawn in chronological order to connect the voltage markings sequentially, and the lines connecting the voltage markings and the voltage markings are highlighted in yellow. Straight lines are drawn in chronological order to connect the bearing temperature markings sequentially, and the lines connecting the bearing temperature markings and the bearing temperature markings are highlighted in green, resulting in a combined three-Y-axis line graph of current, voltage, and bearing temperature. The current factor, voltage factor, and bearing temperature factor are obtained by using a combined three-Y-axis piecewise linear plot of current-voltage-bearing temperature. The weighted average of these factors is calculated to obtain the health assessment factor for the wind turbine equipment. The health status of the wind turbine equipment is then evaluated based on this health assessment factor and the combined three-Y-axis piecewise linear plot of current-voltage-bearing temperature. The process of obtaining the first assessment state of the wind turbine equipment includes: Set current threshold, voltage threshold and bearing temperature threshold. The current threshold, voltage threshold and bearing temperature threshold can be set by experimental data analysis or experience. When the current of the internal circuit of the fan equipment corresponding to the current mark point is greater than or equal to the current threshold, or the voltage of the internal circuit of the fan equipment corresponding to the voltage mark point is greater than or equal to the voltage threshold, or the bearing temperature of the fan equipment corresponding to the bearing temperature mark point is greater than or equal to the bearing temperature threshold, the first evaluation state of the fan equipment is unhealthy. When the current in the internal circuit of the fan equipment corresponding to the current mark point is less than the current threshold, the voltage in the internal circuit of the fan equipment corresponding to the voltage mark point is less than the voltage threshold, and the bearing temperature of the fan equipment corresponding to the bearing temperature mark point is less than the bearing temperature threshold, starting from the second monitoring time point, draw a perpendicular line to the horizontal axis through the previous current mark point, the previous voltage mark point, and the previous bearing temperature mark point, and record it as the first perpendicular line. Draw a parallel line to the horizontal axis through the current mark point and record it as the current parallel line. Draw a parallel line to the horizontal axis through the voltage mark point and record it as the voltage parallel line. Draw a parallel line to the horizontal axis through the bearing temperature mark point and record it as the bearing temperature parallel line. Record the intersection of the current parallel line and the first perpendicular line as the current intersection point. Record the intersection of the voltage parallel line and the first perpendicular line as the voltage intersection point. Record the intersection of the bearing temperature parallel line and the first perpendicular line as the bearing temperature intersection point. Obtain the area of ​​the figure bounded by the straight line between the current marker point and the current intersection point, the straight line between the current marker point and the previous current marker point, and the straight line between the previous current marker point and the current intersection point, and denot it as the current factor. Obtain the area of ​​the figure bounded by the straight line between the voltage marker point and the voltage intersection point, the straight line between the voltage marker point and the previous voltage marker point, and the straight line between the previous voltage marker point and the voltage intersection point, and denot it as the voltage factor; Obtain the area of ​​the figure enclosed by the straight line between the bearing temperature mark point and the bearing temperature intersection point, the straight line between the bearing temperature mark point and the previous bearing temperature mark point, and the straight line between the previous bearing temperature mark point and the bearing temperature intersection point, and record it as the bearing temperature factor. Obtain the weighted average of the current factor, voltage factor, and bearing temperature factor. The weights of the current factor, voltage factor, and bearing temperature factor can be set through experimental data analysis or experience. Use the weighted average of the current factor, voltage factor, and bearing temperature factor as the health assessment factor of the fan equipment at the monitoring time point until the end of the last monitoring time point. The process of obtaining the weighted average of the current factor, voltage factor, and bearing temperature factor includes: The weighted average of the current factor, voltage factor, and bearing temperature factor is denoted as... ; in, ; The weights for the current factor, As the weight of the voltage factor, The weights for the bearing temperature factor. For current factor, For voltage factor, This refers to the bearing temperature factor. Set the threshold for the health assessment factor of the wind turbine equipment. The threshold for the health assessment factor of the wind turbine equipment can be set through experimental data analysis or experience. If all the health assessment factors of the wind turbine equipment are less than the threshold, the first assessment state of the wind turbine equipment is healthy. If there is a health assessment factor of the wind turbine equipment that is greater than or equal to the threshold, the first assessment state of the wind turbine equipment is unhealthy.

[0022] It should be explained that traditional methods typically set thresholds to determine whether relevant parameters are normal. However, significant changes in the current, voltage, or temperature of wind turbine equipment can indicate potential malfunctions and unhealthy conditions. Furthermore, there are correlations between the current, voltage, and temperature of wind turbine equipment. Analyzing only one parameter may lead to inaccurate assessments of the wind turbine's health status, increasing operating costs and reducing factory profits. Therefore, this invention categorizes wind turbine operating data based on the logical causal relationships between them, obtaining power supply-side and mechanical-side operating data. Correlation exists between these data; joint analysis of correlated data significantly increases the accuracy of wind turbine health status assessment. Furthermore, the health status of wind turbine equipment is assessed based on health evaluation factors and a combined three-Y-axis line graph of current-voltage-bearing temperature. Combining changes in current, voltage, and temperature provides a more comprehensive assessment of the wind turbine's health status, thereby improving the accuracy of health status assessment, reducing operating costs, and increasing factory profits.

[0023] The process of determining whether the bearing amplitude of the fan equipment is normal based on mechanical side operating data and whether the sound signal generated by the fan equipment is normal, and when the bearing amplitude of the fan equipment is normal, obtaining the time required for the bearing amplitude to change from normal to abnormal, and performing fan equipment maintenance in advance based on the obtained time includes: The second assessment status includes healthy status and unhealthy status; Under different dust concentration conditions (dust concentration refers to the dust concentration in the air inside the fan equipment), the fan equipment is simulated and operated. Every p minutes, the mass of dust adsorbed by the impeller and the amplitude of the bearing are recorded. The value of p can be set through experimental data analysis or experience. The mass of dust adsorbed by the impeller is fitted with time to obtain a fitting function corresponding to the mass of dust adsorbed by the impeller and time, which is denoted as the mass-time fitting function. The independent variable of the mass-time fitting function is time, and the dependent variable is the mass of dust adsorbed by the impeller. Based on the recorded mass of dust adsorbed by the impeller and the amplitude of the bearing, the mass of dust adsorbed by the impeller and the amplitude of the bearing are fitted to obtain a fitting function corresponding to the mass of dust adsorbed by the impeller and the amplitude of the bearing, which is denoted as the mass-amplitude fitting function. The independent variable of the mass-amplitude fitting function is the mass of dust adsorbed by the impeller, and the dependent variable is the amplitude of the bearing. Set a bearing amplitude threshold. The bearing amplitude threshold can be set through experimental data analysis or experience. When the bearing amplitude of the fan equipment is greater than or equal to the bearing amplitude threshold, the bearing amplitude of the fan equipment is abnormal. When the bearing amplitude of the fan equipment is less than the bearing amplitude threshold, the bearing amplitude of the fan equipment is normal. The dust concentration in the air inside the fan equipment is obtained and recorded as the real-time dust concentration. The bearing amplitude threshold is input into the mass-amplitude fitting function corresponding to the real-time dust concentration to obtain the dust mass adsorbed by the impeller of the fan equipment corresponding to the bearing amplitude threshold, and recorded as the maximum dust mass. The maximum dust mass is input into the mass-time fitting function corresponding to the real-time dust concentration to obtain the time corresponding to the maximum dust mass. The bearing amplitude of the fan equipment is input into the mass-amplitude fitting function corresponding to the real-time dust concentration to obtain the dust mass adsorbed by the impeller of the fan equipment corresponding to the bearing amplitude of the fan equipment, and recorded as the real-time dust mass. The real-time dust mass is then input into the mass-time fitting function corresponding to the real-time dust concentration to obtain the time corresponding to the real-time dust mass. Subtract the time corresponding to the maximum dust mass from the time corresponding to the real-time dust mass to obtain the time required for the bearing amplitude of the fan equipment to reach the bearing amplitude threshold. q minutes before the bearing amplitude of the fan equipment reaches the bearing amplitude threshold, the value of q can be set through experimental data analysis or experience. Then, the fan equipment is shut down and the dust adsorbed by the impeller of the fan equipment is cleaned. It needs to be explained that dust exists inside the fan equipment. As the fan operates for an extended period, this dust will unevenly adhere to the impeller, gradually disrupting the dynamic balance of the fan and increasing the bearing amplitude. Traditional methods typically measure this bearing amplitude, stopping the fan and cleaning the dust when it reaches the maximum permissible value. However, when the bearing amplitude reaches the maximum permissible value, the fan is already in a dangerous state. Failure to stop it in advance could damage the fan and cause a safety accident. Traditional methods cannot provide real-time monitoring. The current wind turbine equipment has a bearing amplitude that takes a certain amount of time to reach its maximum allowable value. This prevents the turbine from being stopped in advance for dust cleaning, increasing the risk of turbine failure and damage, and raising the probability of safety accidents. Therefore, this invention obtains the time required for the bearing amplitude to change from normal to abnormal, and performs wind turbine maintenance in advance based on the obtained time. By monitoring the time required for the bearing amplitude to change from normal to abnormal in real time, the wind turbine equipment can be stopped and cleaned before it is in a dangerous state, thereby reducing the risk of turbine failure and damage, and lowering the probability of safety accidents. The sound signal generated during the operation of the wind turbine is converted to the frequency domain by using Fast Fourier Transform, and the spectrum diagram of the sound signal generated during the operation of the wind turbine is obtained. The spectrum of the sound signal generated during the operation of the wind turbine is input into the wind turbine fault prediction model based on the convolutional neural network. The output of the wind turbine fault prediction model is obtained. The input of the wind turbine fault prediction model is the spectrum, and the output is whether the sound signal generated by the wind turbine is normal or abnormal. When the output of the wind turbine fault prediction model is that the sound signal generated by the wind turbine is abnormal, the reason for the abnormal sound signal generated by the wind turbine is also output. The process of determining whether the wind turbine equipment is in a healthy state based on the first and second assessment states, according to the assessment results, includes: If the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is normal and the bearing amplitude of the wind turbine equipment is normal, then the second evaluation state of the wind turbine equipment is a healthy state. If the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is abnormal or the bearing amplitude of the wind turbine equipment is abnormal, then the second evaluation state of the wind turbine equipment is an unhealthy state. When both the first and second assessment states of the wind turbine equipment are in a healthy state, the wind turbine equipment is in a healthy state. When either the first or second assessment state of the wind turbine equipment is in an unhealthy state, the wind turbine equipment is not in a healthy state. The process of using a pre-defined wind turbine equipment fault prediction model based on a convolutional neural network includes: Based on a convolutional neural network, the loss function of the input layer, convolutional layer, pooling layer, fully connected layer, output layer, and wind turbine equipment fault prediction model is set. The input of the input layer is a spectrum graph, and the output of the output layer is either the sound signal generated by the wind turbine equipment is normal or the sound signal generated by the wind turbine equipment is abnormal. When the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is abnormal, the reason for the abnormal sound signal generated by the wind turbine equipment is also output. Training and using wind turbine equipment fault prediction models specifically includes: Step B1: Obtain r_e samples to form a sample set, and divide the sample set into a training set and a validation set in an 8:2 ratio. Input the training set into the wind turbine equipment fault prediction model for forward propagation. Step B2: Obtain the output results of the wind turbine equipment fault prediction model, calculate the loss value using the loss function, calculate each parameter in the model using the backpropagation algorithm, and update the parameters using the gradient descent algorithm; Step B3: Repeat B1 and B2 until the loss function value of the wind turbine equipment failure prediction model no longer changes. Then import the validation set for validation. If the validation is successful, the trained wind turbine equipment failure prediction model is obtained. If the validation fails, repeat B3. Step B4: Input the spectrum diagram of the sound signal generated during the operation of the wind turbine equipment into the trained wind turbine equipment fault prediction model to obtain the output result of the wind turbine equipment fault prediction model; The process of obtaining r_e samples to form a sample set includes: Obtain the historical spectrum diagrams corresponding to the sound signals generated during the historical operation of r_e wind turbine devices. If the sound signal corresponding to the historical spectrum diagram is abnormal, mark it as an abnormal sound signal generated by the wind turbine device and mark the reason for the abnormal sound signal. If the sound signal corresponding to the historical spectrum diagram is normal, mark it as a normal sound signal generated by the wind turbine device. Take a historical spectrum diagram and its corresponding mark as a sample, and collect r_e samples to form a sample set.

[0024] In this embodiment, the operating data of the wind turbine equipment is classified according to the logical causal relationship between them, obtaining power supply side operating data and mechanical side operating data. Correlation exists between the operating data of the wind turbine equipment; joint analysis of the correlated operating data significantly increases the accuracy of the wind turbine equipment health status assessment, thereby reducing factory operating costs and increasing factory profits. The health status of the wind turbine equipment is assessed based on the wind turbine equipment health evaluation factors and a combined three-Y-axis broken line of current-voltage-bearing temperature. Combining the changes in current, voltage, and temperature of the wind turbine equipment provides a more comprehensive assessment of its health status, thereby improving the accuracy of the wind turbine equipment health status assessment, further reducing factory operating costs and increasing factory profits. When the bearing amplitude of the wind turbine equipment is normal, the time required for the bearing amplitude to change from normal to abnormal is obtained. Based on this time, wind turbine equipment maintenance is performed in advance. Real-time monitoring of the time required for the bearing amplitude to change from normal to abnormal allows for shutdown and dust removal of the wind turbine equipment before it reaches a dangerous state, thereby reducing the risk of wind turbine equipment failure and damage, and lowering the probability of safety accidents.

[0025] Example 2 Please see Figure 2 As shown, for parts not described in detail in this embodiment, please refer to the description in Embodiment 1. A device health status fusion assessment system based on multi-sensor data is provided, including: The data acquisition module is used to acquire the operating status data of the wind turbine equipment during the monitoring period; The data classification module is used to classify the wind turbine equipment operating data according to the logical causal relationship between the operating data of the wind turbine equipment, and obtain the power supply side operating data and the mechanical side operating data. The first assessment module is used to assess the health status of the wind turbine equipment based on the power supply side operating data, and obtain the first assessment status of the wind turbine equipment. The second assessment module is used to assess the health status of the wind turbine equipment based on the mechanical side operating condition data and obtain the second assessment status of the wind turbine equipment. When both the first and second assessment statuses of the wind turbine equipment are in a healthy state, the wind turbine equipment is in a healthy state. When either the first or second assessment status of the wind turbine equipment is in an unhealthy state, the wind turbine equipment is in an unhealthy state.

[0026] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0027] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0028] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

[0029] In conclusion, the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating the health status of equipment based on multi-sensor data fusion, characterized in that, The device health status fusion assessment method based on multi-sensor data includes: Step S1: Obtain wind turbine equipment operating data within the monitoring period using heterogeneous sensors; Step S2: Classify the wind turbine equipment operating data according to the logical causal relationship between the operating data to obtain power supply side operating data and mechanical side operating data; Step S3: Construct a combined three-Y-axis line graph of current-voltage-bearing temperature based on the power supply side operating data. Obtain the current factor, voltage factor, and bearing temperature factor through the combined three-Y-axis line graph of current-voltage-bearing temperature. Calculate the weighted average of the current factor, voltage factor, and bearing temperature factor to obtain the health assessment factor of the wind turbine equipment. Evaluate the health status of the wind turbine equipment based on the health assessment factor of the wind turbine equipment and the combined three-Y-axis line graph of current-voltage-bearing temperature to obtain the first assessment status of the wind turbine equipment. Step S4: Based on the mechanical side operating condition data, determine whether the bearing amplitude of the fan equipment is normal and whether the sound signal generated by the fan equipment is normal. Based on the judgment results, obtain the second evaluation status of the fan equipment. Based on the first and second evaluation statuses of the fan equipment, determine whether the fan equipment is in a healthy state. When the bearing amplitude of the fan equipment is normal, obtain the time required for the bearing amplitude to change from normal to abnormal, and perform fan equipment maintenance in advance based on the obtained time.

2. The device health status fusion assessment method based on multi-sensor data according to claim 1, characterized in that, The operating data of the wind turbine equipment includes the current and voltage of the internal circuits of the wind turbine equipment, the bearing amplitude and bearing temperature of the wind turbine equipment, and the sound signals generated during the operation of the wind turbine equipment. The power supply side operating data includes the current and voltage of the internal circuits of the wind turbine equipment, as well as the bearing temperature of the wind turbine equipment. The mechanical side operating data includes the bearing amplitude of the wind turbine equipment and the sound signals generated during the operation of the wind turbine equipment.

3. The device health status fusion assessment method based on multi-sensor data according to claim 2, characterized in that, The method for constructing a combined current-voltage-bearing temperature three-Y-axis line graph based on power supply side operating condition data includes: The monitoring period is evenly divided into u monitoring time points, and the current, voltage of the internal circuits of the wind turbine equipment and the bearing temperature of the wind turbine equipment are obtained at each monitoring time point. Establish a blank three-Y coordinate system. Set the horizontal axis of the blank three-Y coordinate system to time, set the first Y axis of the blank three-Y coordinate system to current, set the second Y axis of the blank three-Y coordinate system to voltage, and set the third Y axis of the blank three-Y coordinate system to temperature. The current, voltage, and bearing temperature of the internal circuits of the wind turbine equipment at each monitoring time point are filled into a blank three-Y-axis coordinate system. The current, voltage, and bearing temperature markers at each monitoring time point are obtained respectively. Straight lines are drawn in chronological order to connect the current markers, voltage markers, and bearing temperature markers in chronological order, thus obtaining a combined three-Y-axis polygonal graph of current-voltage-bearing temperature.

4. The device health status fusion assessment method based on multi-sensor data according to claim 3, characterized in that, The method for assessing the health status of wind turbine equipment includes: The first assessment status includes a healthy state and an unhealthy state; Set current threshold, voltage threshold and bearing temperature threshold. When the current in the internal circuit of the fan equipment corresponding to the current mark point is greater than or equal to the current threshold, or the voltage in the internal circuit of the fan equipment corresponding to the voltage mark point is greater than or equal to the voltage threshold, or the bearing temperature of the fan equipment corresponding to the bearing temperature mark point is greater than or equal to the bearing temperature threshold, the first evaluation state of the fan equipment is unhealthy. When the current in the internal circuit of the fan equipment corresponding to the current mark point is less than the current threshold, the voltage in the internal circuit of the fan equipment corresponding to the voltage mark point is less than the voltage threshold, and the bearing temperature of the fan equipment corresponding to the bearing temperature mark point is less than the bearing temperature threshold, the health assessment factor of the fan equipment is obtained. Set a threshold for the health assessment factor of the wind turbine equipment. If all the health assessment factors of the wind turbine equipment are less than the threshold, the first assessment status of the wind turbine equipment is a healthy state. If there is a health assessment factor for the wind turbine equipment that is greater than or equal to the threshold of the health assessment factor, then the first assessment state of the wind turbine equipment is an unhealthy state.

5. The device health status fusion assessment method based on multi-sensor data according to claim 4, characterized in that, The method for obtaining the health assessment factors of the wind turbine equipment includes: Starting from the second monitoring time point, draw a perpendicular line to the horizontal axis through the previous current marker, the previous voltage marker, and the previous bearing temperature marker, and record it as the first perpendicular line; Draw a line parallel to the horizontal axis through the current marking point and record it as the current parallel line; draw a line parallel to the horizontal axis through the voltage marking point and record it as the voltage parallel line; draw a line parallel to the horizontal axis through the bearing temperature marking point and record it as the bearing temperature parallel line; record the intersection of the current parallel line and the first perpendicular line as the current intersection point; record the intersection of the voltage parallel line and the first perpendicular line as the voltage intersection point; record the intersection of the bearing temperature parallel line and the first perpendicular line as the bearing temperature intersection point. Obtain the area of ​​the figure bounded by the straight line between the current marker point and the current intersection point, the straight line between the current marker point and the previous current marker point, and the straight line between the previous current marker point and the current intersection point, and denot it as the current factor. Obtain the area of ​​the figure bounded by the straight line between the voltage marker point and the voltage intersection point, the straight line between the voltage marker point and the previous voltage marker point, and the straight line between the previous voltage marker point and the voltage intersection point, and denot it as the voltage factor; Obtain the area of ​​the figure enclosed by the straight line between the bearing temperature mark point and the bearing temperature intersection point, the straight line between the bearing temperature mark point and the previous bearing temperature mark point, and the straight line between the previous bearing temperature mark point and the bearing temperature intersection point, and record it as the bearing temperature factor. The weighted average of the current factor, voltage factor, and bearing temperature factor is calculated, and the calculated weighted average is used as the health assessment factor of the wind turbine equipment at the monitoring time point until the end of the last monitoring time point.

6. The device health status fusion assessment method based on multi-sensor data according to claim 5, characterized in that, The methods for determining whether the bearing amplitude of the fan equipment is normal and whether the sound signal generated by the fan equipment is normal based on mechanical side operating condition data include: Set a bearing amplitude threshold. When the bearing amplitude of the fan equipment is greater than or equal to the bearing amplitude threshold, the bearing amplitude of the fan equipment is abnormal. When the bearing amplitude of the fan equipment is less than the bearing amplitude threshold, the bearing amplitude of the fan equipment is normal. The sound signal generated during the operation of the wind turbine is converted to the frequency domain by using Fast Fourier Transform, and the spectrum diagram of the sound signal generated during the operation of the wind turbine is obtained. The spectrum of the sound signal generated during the operation of the wind turbine is input into the wind turbine fault prediction model based on the convolutional neural network to obtain the output result of the wind turbine fault prediction model. The input of the wind turbine fault prediction model is the spectrum, and the output is whether the sound signal generated by the wind turbine is normal or abnormal.

7. The device health status fusion assessment method based on multi-sensor data according to claim 6, characterized in that, The method for obtaining the second evaluation state of the wind turbine equipment based on the judgment result includes: The second assessment status includes a healthy state and an unhealthy state; If the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is normal and the bearing amplitude of the wind turbine equipment is normal, then the second evaluation state of the wind turbine equipment is a healthy state. If the output of the wind turbine equipment fault prediction model is that the sound signal generated by the wind turbine equipment is abnormal or the bearing amplitude of the wind turbine equipment is abnormal, then the second evaluation state of the wind turbine equipment is an unhealthy state.

8. The device health status fusion assessment method based on multi-sensor data according to claim 7, characterized in that, The method for determining whether the wind turbine equipment is in a healthy state based on the first and second assessment states of the wind turbine equipment includes: The wind turbine is in a healthy state when both its first and second assessment states are healthy. The wind turbine is not in a healthy state when either its first or second assessment state is unhealthy.

9. The device health status fusion assessment method based on multi-sensor data according to claim 8, characterized in that, The method for obtaining the time required for the bearing amplitude to change from normal to abnormal when the bearing amplitude of the wind turbine equipment is normal, and for performing wind turbine equipment maintenance in advance based on the obtained time, includes: Under different dust concentration conditions, the dust concentration refers to the dust concentration in the air inside the fan equipment. The fan equipment is simulated and the mass of dust adsorbed by the impeller and the bearing amplitude of the fan equipment are recorded every p minutes. The mass of dust adsorbed by the impeller of the fan equipment is fitted with time to obtain the fitting function corresponding to the mass of dust adsorbed by the impeller of the fan equipment with time, and is denoted as the mass-time fitting function. The mass of dust adsorbed by the impeller and the bearing amplitude of the fan equipment are fitted to obtain the fitting function corresponding to the mass of dust adsorbed by the impeller and the bearing amplitude of the fan equipment, and are denoted as the mass-amplitude fitting function. The time required for the bearing amplitude of the fan equipment to reach the bearing amplitude threshold is obtained based on the mass-time fitting function and the mass-amplitude fitting function. The fan equipment is stopped q minutes before the bearing amplitude reaches the bearing amplitude threshold, and the dust adsorbed by the impeller of the fan equipment is cleaned.

10. The device health status fusion assessment method based on multi-sensor data according to claim 9, characterized in that, The method for obtaining the time required for the bearing amplitude of the wind turbine to reach the bearing amplitude threshold based on the mass-time fitting function and the mass-amplitude fitting function includes: The dust concentration in the air inside the fan equipment is obtained and recorded as the real-time dust concentration. The bearing amplitude threshold is input into the mass-amplitude fitting function corresponding to the real-time dust concentration to obtain the dust mass adsorbed by the impeller of the fan equipment corresponding to the bearing amplitude threshold, which is recorded as the maximum dust mass. The maximum dust mass is input into the mass-time fitting function corresponding to the real-time dust concentration to obtain the time corresponding to the maximum dust mass. The bearing amplitude of the fan equipment is input into the mass-amplitude fitting function corresponding to the real-time dust concentration to obtain the dust mass adsorbed by the impeller of the fan equipment corresponding to the bearing amplitude of the fan equipment, and recorded as the real-time dust mass. The real-time dust mass is then input into the mass-time fitting function corresponding to the real-time dust concentration to obtain the time corresponding to the real-time dust mass. Subtract the time corresponding to the real-time dust mass from the time corresponding to the maximum dust mass to obtain the time required for the bearing amplitude of the fan equipment to reach the bearing amplitude threshold.

Images