Fan system resonance detection method and intelligent electric appliance

Abstract

The invention relates to a fan system resonance detection method and an intelligent electric appliance. The fan system resonance detection method comprises the steps that a plurality of sensors are arranged on a fan system; constructing a simulation model of the fan system, importing the vibration signals collected by each A-type sensor into the simulation model, and obtaining vibration data of the whole fan system by using a modal extension method; establishing a digital twinborn model, and importing the simulation model, vibration data of the whole fan system and vibration signals acquired by each A-type sensor into the digital twinborn model to obtain a result output by the digital twinborn model; comparing a result output by the digital twin model with a detection result of the B-type sensor, and updating the simulation model; and judging whether the current fan system resonates or not according to the ratio of the maximum amplitude Smax of all the vibration signals of the fan system to S. According to the method, the resonance problem of the fan system can be recognized in real time, then adjustment measures are triggered, abnormal noise is reduced, and resonance generation nodes are recognized.

Description

Technical Field

[0001] This invention relates to the field of intelligent electrical appliance technology, and in particular to a method for detecting resonance in a fan system and an intelligent electrical appliance thereof. Background Technology

[0002] Currently, resonance analysis of wind turbine systems involves two steps: simulation and testing. However, these two steps are often separate and independent. Simulation is only used in the early stages when there is no physical prototype, for simulation prediction during the numerical modeling phase. Testing is only used in the later stages, when a prototype is available, to test and verify the physical prototype. Then, the data from the simulation and testing are compared to check the similarity between the two. The simulation and testing data are independent of each other; test data is not imported into the simulation model to adjust the simulation model parameters.

[0003] Traditional vibration control technologies rely on preset models or manual adjustments, and are completed during the project design phase, making it impossible to adjust parameters in real time based on the actual operating status of the equipment. Furthermore, over time, factors such as winding aging and bearing wear in the motor of the wind turbine system may cause the output excitation to deviate from the settings; the structure may also change its modal information due to loose connections, fatigue, etc., resulting in significant differences from the original design. This makes it difficult to guarantee the accuracy of both simulation and testing. In such cases, identifying resonance and locating the resonance area to facilitate repairs by users or engineers becomes challenging. Summary of the Invention

[0004] The first technical problem to be solved by the present invention is to provide a resonance detection method for wind turbine systems that can identify resonance problems, in contrast to the above-mentioned prior art.

[0005] The second technical problem to be solved by the present invention is to provide an intelligent electrical appliance that applies the above-mentioned wind turbine system resonance detection method.

[0006] The technical solution adopted by the present invention to solve the first technical problem mentioned above is: a method for detecting resonance in a wind turbine system, characterized by comprising the following steps:

[0007] Step 1: Set up multiple sensors on the fan system to be tested to collect vibration signals. The sensors are divided into Class A sensors and Class B sensors that are the same as Class A sensors. Class A sensors are used for modal expansion, and Class B sensors are used for result verification.

[0008] Step 2: Construct a simulation model of the wind turbine system, import the vibration signals collected by each Class A sensor into the simulation model, and use the modal extension method to obtain the vibration data of the entire wind turbine system;

[0009] Step 3: Establish a digital twin model and import the simulation model, the vibration data of the entire wind turbine system, and the vibration signals collected by each Class A sensor into the digital twin model to obtain the output results of the digital twin model;

[0010] Step 4: Compare the results output by the digital twin model with the detection results of the Class B sensor to update the simulation model;

[0011] Step 5: Control the speed of the fan system within the range [V-V0, V+V1], where V is the fixed speed of the fan system at the set gear, and V0 and V1 are preset constants. Use Class A sensors to collect the vibration signals of the current fan system. Import the vibration signals collected by each Class A sensor into the updated simulation model. Additionally, use the modal extension method to obtain the vibration signals of each node on the entire fan system, and filter out the maximum amplitude S of all vibration signals when the fan system speed varies within the range [V-V0, V+V1]. max ;

[0012] Step 6: Determine S max Is the ratio of S greater than S represents the vibration signal amplitude obtained when the fan system operates at a fixed speed V. If it is, it is determined that the current fan system is resonating, and the process proceeds to step 7; otherwise, it is determined that the current fan system is not resonating, and the process ends.

[0013] Step 7: Determine the spatial location of resonance based on the vibration data of the entire wind turbine system reconstructed using the modal extension method in Step 2. If the amplitude of some nodes in the vibration data of the entire wind turbine system is significantly higher than that of neighboring nodes or exceeds the preset upper limit threshold, then the node is determined to be a resonance point.

[0014] Preferably, the simulation model employs a neural network, using the vibration signals collected by each Class A sensor as the input to the neural network, and the material parameters to be optimized in the wind turbine system and the weld point information as the output of the neural network.

[0015] Preferably, the process of updating the simulation model in step 4 is as follows:

[0016] The vibration data of the entire wind turbine system is obtained by using the modal expansion method based on the output of the neural network. The predicted values ​​corresponding to the measurement positions of the Class B sensors are selected from the vibration data of the entire wind turbine system obtained by modal expansion. The loss function is calculated based on the measured value and the predicted value of each Class B sensor, and the output of the neural network is updated using the loss function.

[0017] Preferably, the sensor includes at least one of an accelerometer, a velocity sensor, and a displacement sensor.

[0018] Preferably, the material parameters to be optimized in the wind turbine system are the density, elastic modulus, and Poisson's ratio of the material to be optimized.

[0019] Preferably, the welding point information of the fan system includes the welding point location and the number of welding points.

[0020] Preferably, the number of Class A sensors is at least 5.

[0021] Preferably, the number of the type B sensors is at least three.

[0022] The technical solution adopted by the present invention to solve the second technical problem mentioned above is: a smart appliance, characterized in that: it applies the above-mentioned fan system resonance detection method, wherein the smart appliance is a range hood.

[0023] Compared with existing technologies, the advantages of this invention are: by combining modal extension and digital twin technologies, a virtual wind turbine system model is constructed, thereby optimizing the parameters of the wind turbine system model to identify resonance problems in the wind turbine system in real time, triggering adjustment measures to reduce abnormal noise and facilitate subsequent maintenance by users and engineers. Attached Figure Description

[0024] Figure 1 This is a flowchart of the resonance detection method for a wind turbine system in an embodiment of the present invention. Detailed Implementation

[0025] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0026] like Figure 1 As shown, the wind turbine system resonance detection method in this embodiment includes the following steps:

[0027] Step 1: Set up multiple sensors on the fan system to be tested to collect vibration signals. The sensors are divided into Class A sensors and Class B sensors that are the same as Class A sensors. Class A sensors are used for modal expansion, and Class B sensors are used for result verification.

[0028] The fan system can use a centrifugal fan from a range hood, the structure of which is existing technology and will not be described in detail here; the sensors in this embodiment include at least one of an accelerometer, a velocity sensor, and a displacement sensor; to ensure the reliability of the results, the number of Class A sensors is at least 5; to ensure the accuracy of the digital twin model, the number of Class B sensors is at least 3.

[0029] Step 2: Construct a simulation model of the wind turbine system, import the vibration signals collected by each Class A sensor into the simulation model, and use the modal extension method to obtain the vibration data of the entire wind turbine system;

[0030] The modal expansion method in this embodiment is existing technology and will not be described in detail here.

[0031] Step 3: Establish a digital twin model and import the simulation model, the vibration data of the entire wind turbine system, and the vibration signals collected by each Class A sensor into the digital twin model to obtain the output results of the digital twin model;

[0032] Step 4: Compare the results output by the digital twin model with the detection results of the Class B sensor to update the simulation model;

[0033] Step 5: Control the speed of the fan system within the range [V-V0, V+V1], where V is the fixed speed of the fan system at the set gear, and V0 and V1 are preset constants. Use Class A sensors to collect the vibration signals of the current fan system. Import the vibration signals collected by each Class A sensor into the updated simulation model. Additionally, use the modal extension method to obtain the vibration signals of each node on the entire fan system, and filter out the maximum amplitude S of all vibration signals when the fan system speed varies within the range [V-V0, V+V1]. max ;

[0034] Step 6: Determine S max Is the ratio of S greater than S represents the vibration signal amplitude obtained when the fan system operates at a fixed speed V. If it is, it is determined that the current fan system is resonating, and the process proceeds to step 7; otherwise, it is determined that the current fan system is not resonating, and the process ends.

[0035] Step 7: Based on the vibration data of the entire wind turbine system reconstructed using the modal extension method in Step 2, determine the spatial location of the resonance. If the amplitude of certain nodes in the vibration data of the entire wind turbine system is significantly higher than that of adjacent nodes or exceeds a preset upper limit threshold, then that node is determined to be a resonance point. Engineers can use this information to locate abnormal parts and achieve rapid diagnosis and maintenance of critical wind turbine structures.

[0036] The simulation model described above employs a neural network, using the vibration signals collected by each Class A sensor as input and the material parameters to be optimized in the wind turbine system and the weld point information as output. In this embodiment, the material parameters to be optimized in the wind turbine system are the density, elastic modulus, and Poisson's ratio of the material; the weld point information of the wind turbine system includes the weld point location and the number of weld points.

[0037] The process of updating the simulation model in step 4 is as follows: the vibration data of the entire wind turbine system is obtained by using the modal expansion method to obtain the results of the neural network output. The predicted values ​​corresponding to the measurement positions of the Class B sensors are selected from the vibration data of the entire wind turbine system obtained by modal expansion. The loss function is calculated based on the measured values ​​and predicted values ​​of each Class B sensor, and the results of the neural network output are updated using the loss function.

[0038] In this embodiment, the speed is increased uniformly within 30 rpm before and after the set speed. That is, assuming that the strong speed of the fan system is 1000 rpm, the program quickly increases to 970 rpm, then increases uniformly to 1030 rpm, and then decreases to 1000 rpm. The speed-amplitude curves of all nodes in the speed range of 970-1000 rpm can be obtained, and the node with the largest amplitude is selected.

[0039] This embodiment also relates to a smart appliance that uses the above-mentioned fan system resonance detection method, and the smart appliance is preferably a range hood.

[0040] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for detecting resonance in a wind turbine system, characterized in that... Includes the following steps: Step 1: Set up multiple sensors on the fan system to be tested to collect vibration signals. The sensors are divided into Class A sensors and Class B sensors that are the same as Class A sensors. Class A sensors are used for modal expansion, and Class B sensors are used for result verification. Step 2: Construct a simulation model of the wind turbine system, import the vibration signals collected by each Class A sensor into the simulation model, and use the modal extension method to obtain the vibration data of the entire wind turbine system; Step 3: Establish a digital twin model and import the simulation model, the vibration data of the entire wind turbine system, and the vibration signals collected by each Class A sensor into the digital twin model to obtain the output results of the digital twin model; Step 4: Compare the results output by the digital twin model with the detection results of the Class B sensor to update the simulation model; Step 5: Control the speed of the fan system within the range [V-V0, V+V1], where V is the fixed speed of the fan system at the set gear, and V0 and V1 are preset constants. Use Class A sensors to collect the vibration signals of the current fan system. Import the vibration signals collected by each Class A sensor into the updated simulation model. Additionally, use the modal extension method to obtain the vibration signals of each node on the entire fan system, and filter out the maximum amplitude S of all vibration signals when the fan system speed varies within the range [V-V0, V+V1]. max ; Step 6: Determine S max Is the ratio of S greater than S represents the vibration signal amplitude obtained when the fan system operates at a fixed speed V. If it is, it is determined that the current fan system is resonating, and the process proceeds to step 7; otherwise, it is determined that the current fan system is not resonating, and the process ends. Step 7: Determine the spatial location of resonance based on the vibration data of the entire wind turbine system reconstructed using the modal extension method in Step 2. If the amplitude of some nodes in the vibration data of the entire wind turbine system is significantly higher than that of neighboring nodes or exceeds the preset upper limit threshold, then the node is determined to be a resonance point.

2. The wind turbine system resonance detection method according to claim 1, characterized in that: The simulation model uses a neural network, with the vibration signals collected by each Class A sensor as the input to the neural network, and the material parameters to be optimized in the wind turbine system and the weld point information as the output of the neural network.

3. The method for detecting resonance in a wind turbine system according to claim 2, characterized in that: The process of updating the simulation model in step 4 is as follows: The vibration data of the entire wind turbine system is obtained by using the modal expansion method based on the output of the neural network. The predicted values ​​corresponding to the measurement positions of the Class B sensors are selected from the vibration data of the entire wind turbine system obtained by modal expansion. The loss function is calculated based on the measured value and the predicted value of each Class B sensor, and the output of the neural network is updated using the loss function.

4. The method for detecting resonance in a wind turbine system according to claim 1, characterized in that: The sensor includes at least one of an accelerometer, a velocity sensor, and a displacement sensor.

5. The method for detecting resonance in a wind turbine system according to claim 2, characterized in that: The material parameters to be optimized in the wind turbine system are the density, elastic modulus, and Poisson's ratio of the material to be optimized.

6. The method for detecting resonance in a wind turbine system according to claim 2, characterized in that: The welding point information of the wind turbine system includes the location and number of welding points.

7. The method for detecting resonance in a wind turbine system according to any one of claims 1 to 6, characterized in that: The number of Class A sensors is at least 5.

8. The method for detecting resonance in a wind turbine system according to any one of claims 1 to 6, characterized in that: The number of Class B sensors is at least three.

9. A smart appliance, characterized in that: The application is a fan system resonance detection method as described in any one of claims 1 to 8, wherein the smart appliance is a range hood.

Images