Control method, system and bearing control method for a rotating member

Abstract

The application discloses a kind of control method of rotating member, system and the control method of bearing, wherein, the control method of rotating member includes: obtaining rotating member identification information and rotating state information;When rotating member identification information and the current response rotating member identification information do not match, the initial value of Proportional-Integral-Derivative neural network model is updated according to the obtained rotating member identification information;The control parameter of Proportional-Integral-Derivative controller is calculated using rotating state information and the updated Proportional-Integral-Derivative neural network model;The rotating control value is obtained by processing rotating state information using Proportional-Integral-Derivative controller, to control rotating member.By updating initial value, control parameter is recalculated, and targeted control of different rotating members is realized.The control stability and program universality of rotating member control system are improved.The application solves the technical problem of the decline of rotating member control system stability caused by unreasonable selection of control parameter.

Description

Technical Field

[0001] This invention relates to the field of rotating machinery control, and more specifically, to a control method, system, and bearing control method for rotating components. Background Technology

[0002] In various fields, there is a frequent need for mechanical equipment capable of rotational motion, which drives other mechanisms to rotate; examples include centrifuges, gearboxes, and motors. During the use of such rotating equipment, it is necessary to control the rotational state of the internal rotating components. For instance, in a magnetic levitation motor, the rotational state of the magnetic bearing needs to be controlled to ensure stable output.

[0003] In existing technologies, PID control is commonly used to control rotating components. The PID parameters directly determine the dynamic response speed, control bandwidth, and system stability of the rotating component control system. However, in most existing technologies, the operation of the rotating component control system is maintained by manually setting the PID parameters. However, manually finding the optimal PID parameters is difficult, leading to a decrease in the stability of the rotating component control system. Summary of the Invention

[0004] This application provides a control method, system, and bearing control method for a rotating component, to at least solve the technical problem of decreased stability of the rotating component control system caused by unreasonable selection of PID parameters.

[0005] According to one aspect of the present invention, a method for controlling a rotating component is provided, comprising:

[0006] Obtain rotating component identification information and rotation status information;

[0007] When the rotating component identification information does not match the rotating component identification information of the current response, the initial value of the proportional-integral-differential neural network model is updated according to the acquired rotating component identification information.

[0008] The control parameters of the proportional-integral-derivative (PID) controller are calculated using the rotational state information and the updated PID neural network model.

[0009] The proportional-integral-derivative (PID) controller is updated according to the control parameters, and the updated PID controller is used to process the rotational state information to obtain rotational control values ​​for controlling the rotating component.

[0010] Optionally, the step of calculating the control parameters of the proportional-integral-derivative (PID) controller using the rotational state information and the updated PID neural network model includes:

[0011] The proportional coefficient, integral coefficient, and differential coefficient are calculated based on the displacement deviation information in the rotational state information and the initial value of the proportional-integral-differential neural network model.

[0012] The proportional coefficient, integral coefficient, and derivative coefficient are determined as the control parameters, wherein the control parameters are related to the calculation of the rotation control value.

[0013] Optionally, the displacement deviation information includes the displacement deviation value at the current sampling time, the displacement deviation value at the previous sampling time, and the displacement deviation value at the two previous sampling times.

[0014] Optionally, when the rotating component identification information does not match the rotating component identification information of the current response, the method further includes:

[0015] Calculate the weighting coefficient of the first neuron in the initial value and the preset first bias value to obtain the threshold range of the integral coefficient;

[0016] Calculate the weighting coefficient of the second neuron in the initial value and the preset second bias value to obtain the proportional coefficient threshold range;

[0017] The weighting coefficient of the third neuron in the initial value and the preset third bias value are calculated to obtain the threshold range of the differential coefficient.

[0018] Optionally, after obtaining the rotation control value, the method further includes:

[0019] Map the rotation control value to the target current value;

[0020] The difference between the target current value and the current feedback value is used to obtain the current deviation value;

[0021] The current deviation value is processed using a current controller to obtain the control current;

[0022] The control current is transmitted to the rotating component or an electrical component used to control the rotational state of the rotating component.

[0023] Optionally, after transmitting the control current to the rotating component or an electrical component for controlling the rotational state of the rotating component, the method further includes:

[0024] Obtain the actual position information of the rotating component collected by the displacement sensor in the displacement loop;

[0025] Based on the actual position information of the rotating component and the preset reference position information, the displacement deviation information of the rotating component is calculated.

[0026] The displacement deviation information is processed using the proportional-integral-derivative neural network model and the proportional-integral-derivative controller to obtain the rotation control value;

[0027] The rotation control value is processed using the position adjuster in the displacement loop to obtain the target current value;

[0028] The control current is calculated based on the current feedback value of the coil collected by the current sensor in the current loop and the target current value.

[0029] The control current is used to generate a real-time duty cycle pulse signal and transmit it to the power amplifier to control the coil current;

[0030] Based on the acquisition frequency of the actual position information collected by the displacement sensor, iterative calculations are performed in the dual closed-loop control system composed of the displacement loop and the current loop to control the rotating component by controlling the coil current.

[0031] Optionally, the rotating component is a magnetic levitation bearing.

[0032] Optionally, updating the initial value of the proportional-integral-differential neural network model based on the acquired rotating component identification information includes:

[0033] The identification information of the rotating component is uploaded to the host computer.

[0034] Obtain the initial value sent by the host computer, wherein the initial value is preset in the host computer and corresponds to the identification information of the rotating part, and the initial value refers to the neuron weighting coefficient of the proportional-integral-differential neural network model;

[0035] Replace the original initial values ​​in the proportional-integral-differential neural network model with the initial values ​​issued by the host computer.

[0036] Optionally, the rotating component identification information includes rotating component volume information and rotating component mass information;

[0037] After acquiring the rotating component identification information, the method further includes:

[0038] Determine whether the volume information of the rotating component is the same as the volume information of the rotating component in the currently responded rotating component identification information;

[0039] Determine whether the rotating component mass information is the same as the rotating component mass information in the currently responded rotating component identification information;

[0040] If they are the same, it is determined that the rotating component identification information matches the rotating component identification information of the current response; otherwise, it is determined that the rotating component identification information does not match the rotating component identification information of the current response.

[0041] According to another aspect of the present invention, a control system for a rotating component is also provided, comprising:

[0042] Proportional-integral-differential neural network model, proportional-integral-differential controller, current controller, power amplifier, coil, rotating component, current sensor and displacement sensor;

[0043] The proportional-integral-derivative neural network model is connected to the proportional-integral-derivative controller and is used to calculate the control parameters of the proportional-integral-derivative controller.

[0044] The proportional-integral-derivative (PID) controller is connected to the current controller and the displacement sensor, and is used to replace the original control parameters with the control parameters transmitted by the PID neural network model, and calculate the rotation control value based on the displacement feedback value transmitted by the displacement sensor.

[0045] The current controller is connected to the power amplifier and is used to generate a pulse signal based on the rotation control value;

[0046] The power amplifier is connected to the coil and is used to generate a control current according to the pulse signal and transmit the control current to the coil;

[0047] The coil is connected to the rotating component and is used to control the rotation state of the rotating component according to the control current;

[0048] The current sensor is used to collect the current in the coil and generate a feedback current value;

[0049] The displacement sensor is used to acquire the position of the rotating component and generate displacement feedback values;

[0050] The control system of the rotating component employs the control method described above when controlling the rotating component.

[0051] According to another aspect of the present invention, a storage medium is also provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to perform the control method of the rotating component described above.

[0052] According to another aspect of the present invention, a processor is also provided, the processor being configured to run a program, wherein the program, when running, executes the control method for the rotating component described above.

[0053] According to another aspect of the present invention, a control method for a magnetic levitation bearing is also provided, comprising:

[0054] Obtain bearing identification information and bearing rotation status information;

[0055] When the bearing identification information does not match the bearing identification information of the current response, the initial value of the proportional-integral-differential neural network model is updated according to the acquired bearing identification information.

[0056] The control parameters of the proportional-integral-derivative (PID) controller are calculated using the bearing rotation state information and the updated PID neural network model.

[0057] The proportional-integral-derivative (PID) controller is updated according to the control parameters, and the updated PID controller is used to process the bearing rotation state information to obtain the bearing control value, so as to modify the position of the magnetic levitation bearing.

[0058] In this embodiment of the invention, the initial values ​​of the proportional-integral-derivative (PID) neural network model are replaced. By acquiring the rotating component identification information, it is determined whether the rotating component to be controlled has changed. If a change is confirmed, the initial values ​​of the PID model are updated, allowing the PID model to recalculate the control parameters using the initial values ​​and rotational state information. Finally, the rotational control value is calculated using the updated PID controller to control the rotating component. This achieves the goal of timely updating of control parameters, thereby improving the control stability of the rotating component control system and solving the technical problem of decreased stability in the rotating component control system caused by unreasonable selection of control parameters. Attached Figure Description

[0059] Figure 1 This is a flowchart of an optional control method for a rotating component according to an embodiment of the present invention;

[0060] Figure 2 This is a schematic diagram of an optional control method for a rotating component according to an embodiment of the present invention;

[0061] Figure 3 This is a structural block diagram of an optional control device for a rotating component according to an embodiment of the present invention;

[0062] Figure 4 This is a flowchart of an optional bearing control method according to an embodiment of the present invention;

[0063] Figure 5 This is another flowchart of an optional bearing control method according to an embodiment of the present invention. Detailed Implementation

[0064] The embodiments of the present invention will be described in detail below. To enable those skilled in the art to better understand the present invention, the technical solutions in 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 should fall within the scope of protection of the present invention.

[0065] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0066] According to an embodiment of the present invention, a method for controlling a rotating component is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0067] During operation, rotating equipment requires strict control of rotational parameters to ensure stable rotation of the rotating components. In existing technologies, PID control is commonly used to manage the rotational stability of these components. However, this method typically involves manually setting PID parameters to maintain operation. This approach makes it difficult to find the optimal PID parameters for the currently controlled rotating component, and the PID parameters cannot be adjusted adaptively when the component's model or operating conditions change, leading to a decrease in the stability of the control system.

[0068] To overcome the above-mentioned shortcomings, Figure 1 This is a control method for the rotating component according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0069] Step S101: Obtain the rotating component identification information and rotation status information.

[0070] The rotating component identification information and rotation status information can be acquired simultaneously or separately. For example, the rotating component identification information can be acquired first, and then the rotation status information can be acquired after several steps. That is, this embodiment does not specifically limit the acquisition order and acquisition time of the rotating component identification information and rotation status information.

[0071] In one embodiment, the rotating component identification information refers to information that can identify the type, model, volume, mass, and / or serial number of the rotating component; that is, the difference between two rotating components can be determined based on the rotating component identification information. This difference can be in mass, volume, or both. In another embodiment, the rotating component identification information refers to information that can identify whether a controlled rotating component has been replaced. For example, the rotating component identification information is a low-level signal; after a change in the controlled rotating component, the rotating component control system generates a low-level signal. Upon receiving the low-level signal, the current executing entity can determine that the controlled rotating component has been replaced. Therefore, this embodiment does not specifically limit the structure, form, or content of the rotating component identification information.

[0072] In one embodiment, rotational state information refers to the rotational state of a rotating component. This rotational state information allows for the determination of at least one of the following: angular velocity, linear velocity, acceleration, and position of the rotating component. For example, in one application scenario, the rotational state information is used to characterize the difference between the current rotational position of the rotating component and a preset target position; where "current" refers to the sampling time.

[0073] It should be noted that the rotating component identification information and rotation status information can be acquired in real time, acquired at a certain moment, or previously stored rotating component identification information and rotation status information; this embodiment does not specifically limit this. For example, in one embodiment, the rotating component identification information and rotation status information are acquired in real time, and the rotating component identification information and rotation status information are acquired in groups, that is, one unit of rotating component identification information and rotation status information is obtained at each sampling moment. In another embodiment, the rotating component identification information and rotation status information collected by the sensor are continuously stored in a queue, and the current executing entity retrieves the rotating component identification information and rotation status information from the queue in sequence.

[0074] Step S102: When the rotating part identification information does not match the rotating part identification information of the current response, update the initial value of the proportional-integral-differential neural network model according to the acquired rotating part identification information.

[0075] In one embodiment, the current executing entity acquires the rotating component identification information in real time. When the controlled rotating component changes, the acquired rotating component identification information will differ from the previous one. That is, the rotating component identification information of the current response refers to the identification information of the controlled rotating component before the change. It can also be understood that after the rotating component changes, the rotating component identification information of the rotating component before the change is the current response information. After obtaining the rotating component identification information of the changed rotating component, the entity compares the two to find that the rotating component has changed, and thus determines the changed rotating component as the current rotating component. The corresponding rotating component identification information is then updated to the rotating component identification information of the current response.

[0076] In one embodiment, the mismatch between the rotating component identification information and the currently responded rotating component identification information can be due to completely different content or only partial differences in the content of the information; this embodiment does not impose specific limitations on this. As long as it can be determined that the rotating component identification information does not match the currently responded rotating component identification information after replacing the controlled rotating component, it is acceptable.

[0077] In one embodiment, the rotating component identification information includes an initial value identifier. The corresponding initial value is retrieved from a preset initial value storage table based on the initial value identifier, and then the original initial value in the proportional-integral-differential neural network model is updated to the retrieved initial value. In another embodiment, the rotating component identification information includes information such as the mass and volume of the rotating component. The current executing entity calculates a new initial value using the mass and volume information of the rotating component according to a preset conversion formula, and then updates the original initial value in the proportional-integral-differential neural network model to the retrieved initial value. This embodiment does not specifically limit the process of obtaining the initial value or the method of updating it in the proportional-integral-differential neural network model.

[0078] It should be noted that the proportional-integral-derivative (PID) neural network model is the same as the PID neural network model. This embodiment does not specifically limit the type of PID neural network model, as long as it can calculate the control parameters of the PID controller. For example, it can be a single-neuron PID control neural network.

[0079] In one embodiment, the initial values ​​of the proportional-integral-differential neural network model refer to the weighting coefficients of the neurons. These initial values ​​can be obtained experimentally based on the existing types or models of rotating components and then stored in an initial value table. Each type or model of rotating component corresponds to a set of initial values.

[0080] Step S103: Calculate the control parameters of the proportional-integral-derivative (PID) controller using the rotational state information and the updated PID neural network model.

[0081] In one embodiment, the rotational state information includes information characterizing the rotational state of the rotating component, such as rotational speed and rotational position. The control parameters are obtained by calculating the rotational state information using a proportional-integral-differential neural network model updated from an initial value. Specifically, the control parameters include a proportional coefficient. Integral coefficient and differential coefficients .

[0082] It should be noted that, in one embodiment, the state information of the rotating component is collected in real time, and the current executing entity acquires it in real time. Therefore, the control parameters are also continuously updated with the collection frequency. The proportional-integral-differential neural network model iteratively calculates and updates the control parameters in real time based on the collection frequency of the rotating component's state information.

[0083] Step S104: Update the proportional-integral-derivative (PID) controller according to the control parameters and use the updated PID controller to process the rotational state information to obtain rotational control values ​​for controlling the rotating component.

[0084] In one embodiment, after obtaining the control parameters, the proportional-integral-derivative (PID) controller is updated to replace the original control parameters. Then, the updated PID controller is used to process the rotational state information to obtain the rotational control value.

[0085] Specifically, the rotation control value can be a control value used to control the rotational component to change its position, a control value used to control the rotational component to change its speed, or a control value that changes both position and speed; this embodiment does not impose specific limitations on this. Furthermore, the rotation control value can be directly applied to the rotating component to change its rotational state; it can also be applied to the driving device of the rotating component, such as a motor or coil. Therefore, the rotation control value takes a corresponding form for different types of driving devices for the rotating component. For example, when the driving device is a motor, the rotation control value can be an electromagnetic wave signal used to change the motor's speed, thereby changing the rotational speed of the rotating component; as another example, when the driving device is a coil, the rotation control value is current used to change the current within the coil to control the rotational speed or position of the rotating component.

[0086] Through the above steps, the identification information of the rotating component is used to determine whether the controlled rotating component has been replaced. After the controlled rotating component is replaced, initial values ​​applicable to the currently controlled rotating component can be obtained based on the rotating component identification information. Then, the obtained initial values ​​are used to update the proportional-integral-derivative (PID) neural network model, and the PID model is used to calculate the control parameters of the PID controller, thus updating the PID controller. This makes the rotation control values ​​calculated by the PID controller more suitable for the replaced rotating component, thereby improving the control accuracy and quality of the rotating component, enabling the control system of the rotating component to control the rotating component more stably, and improving the stability and reliability of the rotating component control system.

[0087] Optionally, the step of calculating the control parameters of the proportional-integral-derivative (PID) controller using rotational state information and the updated PID neural network model includes:

[0088] The proportional coefficient, integral coefficient, and derivative coefficient in the control parameters are calculated based on the displacement deviation information in the rotational state information and the initial value of the proportional-integral-differential neural network model.

[0089] The proportional coefficient, integral coefficient, and derivative coefficient are determined as control parameters, which are related to the calculation of the rotational control value.

[0090] In one embodiment, the displacement of the rotating component is monitored, specifically the distance by which the rotating component deviates from the target position during rotation. This displacement is named displacement deviation information and belongs to the rotation state information. In specific calculations, the displacement deviation information can be obtained by subtracting the reference position of the rotating component from the monitored actual position.

[0091] In one embodiment, such as Figure 2 As shown, the neural network PID controller includes a proportional-integral-derivative (PID) neural network model and a PID controller, utilizing displacement feedback values ​​collected by a displacement sensor. and reference position By subtracting, displacement deviation information is obtained. ;Will The proportional coefficient, integral coefficient, and derivative coefficient in the control parameters are calculated as inputs to the proportional-integral-differential neural network model.

[0092] Specifically, in one application scenario, the calculation process is as follows:

[0093] (1)

[0094] (2)

[0095] (3)

[0096] In the formula,

[0097] (4)

[0098] in, , , The learning rate is the integral, proportional, or differential; K is the proportionality coefficient of the neuron. For performance indicators or progressive signals; These are the weighting coefficients for the neurons, which are the initial values ​​for the proportional-integral-differential neural network model. This is the control variable at the current sampling time; This is an adjustable coefficient; This is the difference between the expected output and the actual output at the current sampling moment, i.e., the displacement deviation information; This is the control variable at the current sampling time.

[0099] The output control quantity of a classic incremental PID controller is:

[0100] (5)

[0101] In the formula, , , These are the proportional, integral, and differential coefficients, respectively.

[0102] This leads to the neural network PID control parameters:

[0103] (6)

[0104] In the formula, the subscripts MIN and MAX represent the minimum and maximum values ​​of the parameter range.

[0105] By using the above steps, the proportional coefficient, integral coefficient, and derivative coefficient at the current sampling moment are continuously updated using the initial value and displacement deviation information of the proportional-integral-derivative neural network model. This improves the accuracy of the control quantity calculated by the proportional-integral-derivative controller for the current sampling moment, thereby improving the accuracy of displacement control of the rotating component.

[0106] Optionally, the displacement deviation information includes the displacement deviation value at the current sampling time, the displacement deviation value at the previous sampling time, and the displacement deviation value at the two previous sampling times.

[0107] Wherein, if the displacement deviation value at the current sampling time is k, then the displacement deviation value at the previous sampling time is k-1, and the displacement deviation value at the two previous sampling times is k-2.

[0108] Through the above steps, the displacement deviation information contains displacement deviation values ​​at multiple sampling times, which helps to improve the applicability of the proportional coefficient, integral coefficient, and derivative coefficient corresponding to the current sampling time, enabling the control quantity for the current sampling time to more accurately control the rotation state of the rotating component.

[0109] Optionally, when the rotating component identification information does not match the rotating component identification information of the current response, the method further includes:

[0110] Step S401: Calculate the weighting coefficient of the first neuron in the initial value and the preset first bias value to obtain the threshold range of the integral coefficient;

[0111] After updating the initial values ​​of the proportional-integral-differential neural network model, the threshold range of the integral coefficients is simultaneously updated to calculate a new threshold range for the integral coefficients. Specifically, in one embodiment, a first bias value, bate1, is preset. The weighted coefficients of the first neuron in the initial values ​​are added to bate1 to obtain the upper limit threshold of the threshold range for the integral coefficients. The weighted coefficients of the first neuron in the initial values ​​are subtracted from bate1 to obtain the lower limit threshold of the threshold range for the integral coefficients. The weighted coefficients of the first neuron correspond to the integral coefficients.

[0112] Step S402: Calculate the weighting coefficient of the second neuron in the initial value and the preset second bias value to obtain the proportional coefficient threshold range;

[0113] After updating the initial values ​​of the proportional-integral-differential neural network model, a new proportional coefficient threshold interval is calculated during the simultaneous updating of the proportional coefficient threshold. Specifically, in one embodiment, a second bias value, bate2, is preset. The weighted coefficient of the second neuron in the initial value is added to bate2 to obtain the upper limit threshold of the proportional coefficient threshold interval. The weighted coefficient of the second neuron in the initial value is subtracted from bate2 to obtain the lower limit threshold of the proportional coefficient threshold interval. The weighted coefficient of the second neuron corresponds to the proportional coefficient.

[0114] Step S403: Calculate the weighting coefficient of the third neuron in the initial value and the preset third bias value to obtain the threshold range of the differential coefficient.

[0115] After updating the initial values ​​of the proportional-integral-differential neural network model, a new threshold range for the differential coefficients is calculated during the simultaneous updating of the threshold values ​​of the differential coefficients. Specifically, in one embodiment, a third bias value, bate3, is preset. The weighted coefficients of the third neuron in the initial values ​​are added to bate3 to obtain the upper limit threshold of the differential coefficient threshold range. The weighted coefficients of the third neuron in the initial values ​​are subtracted from bate3 to obtain the lower limit threshold of the differential coefficient threshold range. The weighted coefficients of the third neuron correspond to the differential coefficients.

[0116] Furthermore, it should be noted that the first bias value bate1, the second bias value bate2, and the third bias value bate3 can be the same or different. This embodiment does not limit this, and the specific values ​​are determined based on the parameters of the rotating component and the parameters in actual use.

[0117] Through the above steps, the initial values ​​of the proportional-integral-differential neural network model are updated simultaneously with the threshold range of the control coefficients. Since the updated initial values ​​are adapted to the rotating component in the current response, the updated threshold range helps to ensure that the adapted rotating component can obtain accurate control quantities under various operating conditions, thereby ensuring that the same rotating component can be well controlled under different operating conditions.

[0118] Optionally, after obtaining the rotation control value, the method further includes:

[0119] Step S501: Map the rotation control value to the current target value.

[0120] Specifically, in one embodiment, such as Figure 2 As shown, the rotating component control system is a dual closed-loop system, wherein the rotation control value... The value output from the displacement loop is used as a reference value for the current loop, i.e., as the target current value for the current loop. .

[0121] Step S502: Subtract the target current value from the current feedback value to obtain the current deviation value.

[0122] Among them, such as Figure 2 As shown, the current sensor in the current loop collects the current in the coil in real time to obtain the current feedback value. ;Will - The current deviation value was calculated.

[0123] Step S503: Process the current deviation value using a current controller to obtain the control current.

[0124] Among them, such as Figure 2 As shown, the current deviation value is used as the input of the current controller, and the current controller outputs a real-time duty cycle PWM signal to the power amplifier, so that the power amplifier outputs a control current for controlling the rotating parts.

[0125] Step S504: Transmit the control current to the rotating component or the electrical components used to control the rotation state of the rotating component.

[0126] In one embodiment, such as Figure 2As shown, the rotating component is the rotor of the motor, and the rotational position of the rotor is controlled by a magnetic levitation bearing. By transmitting control current to the coil of the magnetic levitation bearing and changing the current value within the coil, the position of the rotor is controlled, thus stabilizing the rotating component at the target position.

[0127] By using the above steps to change the control current of the rotating component by changing the rotation control value, the rotation state of the rotating component can be changed. This is convenient, quick, and helps to improve control accuracy.

[0128] Optionally, such as Figure 2 As shown, after transmitting control current to the rotating component or an electrical component for controlling the rotational state of the rotating component, the method further includes:

[0129] S601. Obtain the actual position information of the rotating component collected by the displacement sensor in the displacement loop.

[0130] The rotating component is controlled by a dual closed-loop feedback system, including a displacement loop and a current loop. The displacement loop monitors the position of the rotating component, while the current loop monitors and controls the control current of the rotating component; that is, the rotational position of the rotating component is controlled by changing the current value.

[0131] In one embodiment, the actual position information of the rotating component is the coordinate information or distance value information of the rotating component from the reference position when the displacement sensor collects the data.

[0132] S602. Based on the actual position information of the rotating component and the preset reference position information, calculate the displacement deviation information of the rotating component.

[0133] S603. The displacement deviation information is processed using a proportional-integral-derivative neural network model and a proportional-integral-derivative controller to obtain rotation control values.

[0134] S604. The rotation control value is processed by the position adjuster in the displacement loop to obtain the target current value.

[0135] S605. The control current is calculated based on the current feedback value of the coil and the target current value collected by the current sensor in the current loop.

[0136] S606 uses the control current to generate a real-time duty cycle pulse signal and transmits it to the power amplifier to control the coil current.

[0137] S607. Based on the acquisition frequency of the actual position information collected by the displacement sensor, iterative calculations are performed in the dual closed-loop control system composed of the displacement loop and the current loop to control the rotating component by controlling the coil current.

[0138] By employing the above steps and using a dual closed-loop feedback method to control the rotational state of the rotating component, the control accuracy can be improved.

[0139] Optionally, the rotating component is a magnetic levitation bearing.

[0140] Optionally, updating the initial values ​​of the proportional-integral-differential neural network model based on the acquired rotating component identification information includes:

[0141] Step S801: Upload the rotating part identification information to the host computer.

[0142] In one embodiment, since the specific rotating component can be identified through the rotating component identification information, the rotating component identification information is directly uploaded to the host computer, and the host computer can find the initial value of the corresponding rotating component through the rotating component identification information.

[0143] Step S802: Obtain the initial value sent by the host computer.

[0144] The initial values ​​are pre-set in the host computer, and each initial value corresponds to the rotating part identification information. That is, a unique initial value can be found based on the rotating part identification information. The initial value refers to the weighting coefficients of the neurons in the proportional-integral-differential neural network model, i.e., the coefficients in the above formula. .

[0145] Step S803: Replace the corresponding initial values ​​in the proportional-integral-differential neural network model.

[0146] In one embodiment, the proportional-integral-differential neural network model is provided with initial values. These initial values ​​can be the default initial values ​​of the proportional-integral-differential neural network model, or initial values ​​assigned when the model was previously used. The initial values ​​obtained from the host computer are used to replace the initial values, thus updating the proportional-integral-differential neural network model.

[0147] Through the above steps, the host computer has multiple preset initial values, each corresponding to a type, model, or rotating component. When the rotating component changes, the corresponding information is transmitted to the host computer, which then sends out the new initial values. This process is convenient and quick, improving the control efficiency of the rotating component and enabling it to quickly regain stable rotational state while obtaining optimized control parameters.

[0148] Optionally, the rotating component identification information includes rotating component volume information and rotating component mass information.

[0149] After acquiring the rotating component identification information, the method further includes:

[0150] Step S901: Determine whether the volume information of the rotating component is the same as the volume information of the rotating component in the current response rotating component identification information.

[0151] In one embodiment, rotating component identification information is acquired in real time, and this real-time information is compared with the currently responding rotating component identification information. Specifically, the volume information of the rotating components in both cases is compared to determine if they are the same. The volume information of the rotating components allows for a quick determination of whether the rotating component has changed, thereby determining whether the initial values ​​of the proportional-integral-differential neural network model need to be updated.

[0152] Step S902: Determine whether the rotating component mass information is the same as the rotating component mass information in the current response rotating component identification information.

[0153] Similarly, in one embodiment, not only the volume information of the rotating component is compared, but also the mass information of the rotating component is compared to reduce the probability of misjudgment.

[0154] If they are the same, the rotating part identification information is determined to match the rotating part identification information of the current response; otherwise, the rotating part identification information is determined to not match the rotating part identification information of the current response.

[0155] The above steps, using volume and mass to determine whether a rotating component has changed, are convenient, quick, and less prone to errors, helping to reduce computational resource consumption. Furthermore, volume and mass are easy to collect; they can be obtained using sensors with appropriate functions, or calculated from material and physical images. This reduces the tediousness of determining whether the rotating component identification information matches the currently responding rotating component identification information.

[0156] In summary, the control method for a rotating component in this embodiment determines whether the controlled rotating component has changed by acquiring its volume and mass information. If a change has occurred, the method retrieves the initial value corresponding to the changed rotating component from the host computer based on the rotating component identification information. Then, the retrieved initial value is assigned to a proportional-integral-differential neural network (PID) model, which calculates the control parameters. Finally, the control parameters are assigned to a PID controller, enabling the PID controller to use control parameters suitable for the controlled rotating component to calculate high-precision rotation control values, ensuring the rotational state of the rotating component. This solves the technical problem of decreased stability in the rotating component control system caused by unreasonable selection of control parameters.

[0157] This application also provides a control system for a rotating component, including a proportional-integral-differential neural network model, a proportional-integral-differential controller, a current controller, a power amplifier, a coil, a rotating component, a current sensor, and a displacement sensor;

[0158] The proportional-integral-derivative neural network model is connected to the proportional-integral-derivative controller and is used to calculate the control parameters of the proportional-integral-derivative controller.

[0159] The proportional-integral-derivative (PID) controller is connected to the current controller and the displacement sensor, and is used to replace the original control parameters with the control parameters transmitted by the PID neural network model, and calculate the rotation control value based on the displacement feedback value transmitted by the displacement sensor.

[0160] The current controller is connected to the power amplifier and is used to generate a pulse signal based on the rotation control value;

[0161] The power amplifier is connected to the coil and is used to generate a control current according to the pulse signal and transmit the control current to the coil;

[0162] The coil is connected to the rotating component and is used to control the rotation state of the rotating component according to the control current;

[0163] The current sensor is used to collect the current in the coil and generate a feedback current value;

[0164] The displacement sensor is used to acquire the position of the rotating component and generate displacement feedback values;

[0165] The control system of the rotating component employs the control method described above when controlling the rotating component.

[0166] This application also provides a control device for a rotating component, such as... Figure 3 As shown, it includes an acquisition module 1, used to acquire rotating component identification information and rotation status information;

[0167] Initial value module 2 is used to update the initial value of the proportional-integral-differential neural network model according to the acquired rotating part identification information when the rotating part identification information does not match the rotating part identification information of the current response.

[0168] Parameter module 3 is used to calculate the control parameters of the proportional-integral-derivative controller using the rotational state information and the updated proportional-integral-derivative neural network model.

[0169] Control module 4 is used to update the proportional-integral-derivative (PID) controller according to the control parameters and process the rotational state information using the updated PID controller to obtain rotational control values ​​for controlling the rotating component.

[0170] Optionally, the parameter module 3 includes a parameter unit for calculating the proportional coefficient, integral coefficient, and derivative coefficient in the control parameters based on the displacement deviation information in the rotational state information and the initial value of the proportional-integral-differential neural network model.

[0171] Optionally, the displacement deviation information includes the displacement deviation value at the current sampling time, the displacement deviation value at the previous sampling time, and the displacement deviation value at the two previous sampling times.

[0172] Optionally, the device further includes a threshold module for calculating the initial value of the weighting coefficient of the first neuron and a preset first bias value to obtain the integral coefficient threshold range;

[0173] The initial value of the weighting coefficient of the second neuron and the preset second bias value are used to obtain the threshold range of the proportional coefficient.

[0174] The initial value of the weighting coefficient of the third neuron and the preset third bias value are used to obtain the threshold range of the differential coefficient.

[0175] Optionally, the device further includes a target value module for mapping the rotation control value to a target current value;

[0176] The deviation value module is used to calculate the difference between the target current value and the current feedback value to obtain the current deviation value.

[0177] A current module is used to process the current deviation value using a current controller to obtain a control current;

[0178] An output module is used to transmit the control current to the rotating component or an electrical component for controlling the rotational state of the rotating component.

[0179] Optionally, the initial value module 2 includes an uploading unit for uploading the rotating component identification information to a host computer;

[0180] A receiving unit is used to obtain the initial value sent by the host computer, wherein the initial value is preset in the host computer and corresponds to the identification information of the rotating part, and the initial value refers to the neuron weighting coefficient of the proportional-integral-differential neural network model;

[0181] The replacement unit is used to replace the corresponding original initial value in the proportional-integral-differential neural network model.

[0182] Optionally, the rotating component identification information includes rotating component volume information and rotating component mass information;

[0183] The device further includes a judgment module for judging whether the volume information of the rotating component is the same as the volume information of the rotating component in the currently responded rotating component identification information.

[0184] Determine whether the rotating component mass information is the same as the rotating component mass information in the currently responded rotating component identification information;

[0185] If they are the same, it is determined that the rotating component identification information matches the rotating component identification information of the current response; otherwise, it is determined that the rotating component identification information does not match the rotating component identification information of the current response.

[0186] Based on the above, by acquiring the rotating component identification information using module 1, it is possible to determine whether the controlled rotating component has changed; by acquiring the rotation state information, the rotation state of the controlled rotating component can be determined. After the controlled rotating component changes, the initial value module 2 updates the initial value of the proportional-integral-differential neural network model based on the rotating component identification information. Simultaneously with updating the initial value, the threshold range of the control parameters is also updated. Then, the parameter module 3 calculates the control parameters using the proportional-integral-differential neural network model. The control module 4 calculates the rotation control value using the control parameters and the rotation state information. Because the initial value is updated, the control parameters are also updated. The updated control parameters have a higher degree of fit with the currently controlled rotating component, resulting in a more accurate rotation control value, which ensures stable rotation of the controlled rotating component and improves the control stability of the rotating component control system.

[0187] This invention also provides a storage medium, the storage medium including a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to execute the control method for the rotating component described above.

[0188] This invention also provides a processor for running a program, wherein the program executes the control method for the rotating component described above.

[0189] This invention also provides a control method for magnetic levitation bearings, such as... Figure 4 As shown, it includes:

[0190] Step S10: Obtain bearing identification information and bearing rotation status information;

[0191] Step S20: When the bearing identification information does not match the bearing identification information of the current response, update the initial value of the proportional-integral-differential neural network model according to the acquired bearing identification information.

[0192] Step S30: Calculate the control parameters of the proportional-integral-derivative (PID) controller using the bearing rotation state information and the updated PID neural network model;

[0193] Step S40: Update the proportional-integral-derivative (PID) controller according to the control parameters and use the updated PID controller to process the bearing rotation state information to obtain the bearing control value, so as to modify the position of the magnetic levitation bearing.

[0194] For ease of understanding, such as Figure 5As shown, when the bearing identification information does not match the bearing identification information of the current response, it can be determined that the size of the bearing has changed, that is, the controlled bearing has changed; otherwise, it can be determined that the bearing has not changed and is the same bearing.

[0195] After a bearing change, the initial values ​​of the proportional-integral-differential neural network (PIN) model are first updated using a host computer. Then, the PNI model iteratively calculates and outputs control parameters and adjusts their upper and lower thresholds as the sampling time progresses. Finally, the bearing control values ​​are obtained using the control parameters and their upper and lower thresholds, thus changing the bearing's operating state.

[0196] If the bearing control value can stabilize the bearing at the target position, it proves that the current initial value can meet the current bearing operation. Then, the host computer writes the initial value into the EEPROM register controlled by the DSP for solidification.

[0197] After the bearing is changed, a proportional-integral-differential neural network model iteratively calculates and outputs control parameters at each sampling time. Using these control parameters and their upper and lower thresholds, the bearing control value is obtained, thus changing the bearing's operating state. If the bearing control value can stabilize the bearing at the target position, it proves that the currently used initial value is sufficient for the bearing's operation. Subsequently, the initial value is written into the EEPROM register controlled by the DSP and permanently stored via the host computer.

[0198] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0199] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0200] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0201] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0202] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0203] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0204] 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 principle 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 controlling a rotating component, characterized in that, include: Real-time acquisition of rotating component identification information and rotation status information, wherein the rotating component identification information refers to information that can identify whether the controlled rotating component has been replaced, and the rotating component includes a magnetic levitation bearing; When the rotating part identification information does not match the rotating part identification information of the current response, the initial value of the proportional-integral-differential neural network model is updated according to the rotating part identification information acquired in real time. The initial value refers to the neuron weighting coefficient of the proportional-integral-differential neural network model. The control parameters of the proportional-integral-derivative (PID) controller are calculated using the real-time acquired rotational state information and the updated PID neural network model. The proportional-integral-derivative (PID) controller is updated according to the control parameters, and the updated PID controller is used to process the rotational state information to obtain rotational control values ​​for controlling the rotating component. The initial values ​​of the proportional-integral-differential neural network model are updated based on the real-time acquired rotating component identification information, including: The rotating component identification information includes an initial value identifier. The corresponding initial value is retrieved from the preset initial value storage table according to the initial value identifier, and then the original initial value in the proportional-integral-differential neural network model is updated with the retrieved initial value. Alternatively, the rotating component identification information includes the mass and volume of the rotating component. According to the preset conversion formula, a new initial value is calculated using the mass and volume of the rotating component. Then, the original initial value in the proportional-integral-differential neural network model is updated to the retrieved initial value. Alternatively, the rotating component identification information is uploaded to the host computer; the initial value sent by the host computer is obtained, wherein the initial value is preset in the host computer and corresponds to the rotating component identification information; and the original initial value in the proportional-integral-differential neural network model is replaced with the initial value sent by the host computer.

2. The control method for a rotating component according to claim 1, characterized in that, The control parameters of the proportional-integral-derivative (PID) controller are calculated using the rotational state information and the updated PID neural network model, including: The proportional coefficient, integral coefficient, and differential coefficient are calculated based on the displacement deviation information in the rotational state information and the initial value of the proportional-integral-differential neural network model. The proportional coefficient, integral coefficient, and derivative coefficient are determined as the control parameters, wherein the control parameters are related to the calculation of the rotation control value.

3. The control method for a rotating component according to claim 2, characterized in that, The displacement deviation information includes the displacement deviation value at the current sampling time, the displacement deviation value at the previous sampling time, and the displacement deviation value at the two sampling times before that.

4. The control method for a rotating component according to claim 2, characterized in that, When the rotating component identification information does not match the rotating component identification information of the current response, the method further includes: Calculate the weighting coefficient of the first neuron in the initial value and the preset first bias value to obtain the threshold range of the integral coefficient; Calculate the weighting coefficient of the second neuron in the initial value and the preset second bias value to obtain the proportional coefficient threshold range; The weighting coefficient of the third neuron in the initial value and the preset third bias value are calculated to obtain the threshold range of the differential coefficient.

5. The control method for a rotating component according to claim 1, characterized in that, After obtaining the rotation control value, the method further includes: Map the rotation control value to the target current value; The difference between the target current value and the current feedback value is used to obtain the current deviation value; The current deviation value is processed using a current controller to obtain the control current; The control current is transmitted to the rotating component or an electrical component used to control the rotational state of the rotating component.

6. The control method for a rotating component according to claim 5, characterized in that, After transmitting the control current to the rotating component or an electrical component for controlling the rotational state of the rotating component, the method further includes: Obtain the actual position information of the rotating component collected by the displacement sensor in the displacement loop; Based on the actual position information of the rotating component and the preset reference position information, the displacement deviation information of the rotating component is calculated. The displacement deviation information is processed using the proportional-integral-derivative neural network model and the proportional-integral-derivative controller to obtain the rotation control value; The rotation control value is processed using the position adjuster in the displacement loop to obtain the target current value; The control current is calculated based on the current feedback value of the coil collected by the current sensor in the current loop and the target current value. The control current is used to generate a real-time duty cycle pulse signal and transmit it to the power amplifier to control the coil current; Based on the acquisition frequency of the actual position information collected by the displacement sensor, iterative calculations are performed in the dual closed-loop control system composed of the displacement loop and the current loop to control the rotating component by controlling the coil current.

7. The control method for a rotating component according to any one of claims 1-6, characterized in that, The rotating component identification information includes rotating component volume information and rotating component mass information; After acquiring the rotating component identification information, the method further includes: Determine whether the volume information of the rotating component is the same as the volume information of the rotating component in the currently responded rotating component identification information; Determine whether the rotating component mass information is the same as the rotating component mass information in the currently responded rotating component identification information; If they are the same, it is determined that the rotating component identification information matches the rotating component identification information of the current response; otherwise, it is determined that the rotating component identification information does not match the rotating component identification information of the current response.

8. A control system for a rotating component, characterized in that, include: Proportional-integral-differential neural network model, proportional-integral-differential controller, current controller, power amplifier, coil, rotating component, current sensor and displacement sensor; The proportional-integral-derivative neural network model is connected to the proportional-integral-derivative controller and is used to calculate the control parameters of the proportional-integral-derivative controller. The proportional-integral-derivative (PID) controller is connected to the current controller and the displacement sensor, and is used to replace the original control parameters with the control parameters transmitted by the PID neural network model, and calculate the rotation control value based on the displacement feedback value transmitted by the displacement sensor. The current controller is connected to the power amplifier and is used to generate a pulse signal based on the rotation control value; The power amplifier is connected to the coil and is used to generate a control current according to the pulse signal and transmit the control current to the coil; The coil is connected to the rotating component and is used to control the rotation state of the rotating component according to the control current; The current sensor is used to collect the current in the coil and generate a feedback current value; The displacement sensor is used to acquire the position of the rotating component and generate displacement feedback values; The control system of the rotating component employs the control method of the rotating component as described in any one of claims 1-7 when controlling the rotating component.

9. A control method for a magnetic levitation bearing, characterized in that, include: Obtain bearing identification information and bearing rotation status information, wherein the bearing identification information refers to information that can identify whether the controlled bearing has been replaced; When the bearing identification information does not match the bearing identification information of the current response, the initial value of the proportional-integral-differential neural network model is updated according to the acquired bearing identification information, where the initial value refers to the weighting coefficient of the neuron. The control parameters of the proportional-integral-derivative (PID) controller are calculated using the bearing rotation state information and the updated PID neural network model. The proportional-integral-derivative (PID) controller is updated according to the control parameters, and the updated PID controller is used to process the bearing rotation state information to obtain the bearing control value, so as to modify the position of the magnetic levitation bearing. The process of updating the initial values ​​of the proportional-integral-differential neural network model based on the acquired bearing identification information includes: The bearing identification information includes an initial value identifier. The corresponding initial value is retrieved from the preset initial value storage table according to the initial value identifier. Then, the original initial value in the proportional-integral-differential neural network model is updated with the retrieved initial value. Alternatively, the bearing identification information may include the bearing's mass and volume. Based on a preset conversion formula, a new initial value is calculated using the bearing's mass and volume. Then, the original initial value in the proportional-integral-differential neural network model is updated to the retrieved initial value. Alternatively, the bearing identification information is uploaded to a host computer; the initial value sent by the host computer is obtained, wherein the initial value is preset in the host computer and corresponds to the bearing identification information; and the original initial value in the proportional-integral-differential neural network model is replaced with the initial value sent by the host computer.

Images