A monostable electromagnetic actuator and an active vibration isolation control method and system thereof
By designing a monostable electromagnetic actuator and adaptive control method for neural networks, the problem of insufficient dynamic response of existing electromagnetic actuators is solved, and the high-performance effect of high-frequency vibration isolation and low-frequency vibration reduction is achieved, which improves the operation Overall performance of the device.
Patent Information
- Application Number
- CN202210600208.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-05-30
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2042-05-30
AI Technical Summary
The existing electromagnetic actuators have insufficient dynamic response and cannot achieve high-performance vibration suppression of the target structure, especially at high and low frequencies, which have poor vibration isolation.
A monostable electromagnetic actuator is designed, using a structure with elastic boundaries, combined with the neural network adaptive active vibration isolation control method, parameter identification is performed through the first neural network and the second neural network adjusts the controller parameters to achieve the The precise control of the actuator output signal uses elastic boundaries and electromagnetic forces to provide excellent vibration isolation performance at high frequencies, and generates an inverted output force by actively controlling the driving motion at low frequencies.
It achieves efficient power output performance, can provide excellent vibration isolation performance at high frequencies, and achieves high-performance vibration damping effect at low frequencies through active control, which significantly improves the dynamic response and control accuracy of the actuator.
Smart Images

Figure CN114944741B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of active vibration isolation control, and in particular relates to a monostable electromagnetic actuator and an active vibration isolation control method and system thereof. Background Art
[0002] Actuators, also known as exciters, can exert control force on the controlled object according to a certain control law. With the development of active vibration control technology, the requirements for actuators are getting higher and higher. Currently, electromagnetic actuators, hydraulic actuators, piezoelectric actuators, magnetostrictive actuators, electrostrictive actuators, pneumatic actuators, etc. are commonly used. Electromagnetic actuators are mainly suitable for occasions with high frequencies and not too large control forces.
[0003] Existing electromagnetic actuators usually use inertial electromagnetic actuators. However, inertial electromagnetic actuators have disadvantages such as small electromagnetic force coefficient and small actuator output force. Other common electromagnetic actuators also have problems such as complex structure and insufficient dynamic response, and cannot achieve high-performance vibration suppression of the target structure. Summary of the invention
[0004] In view of this, the present invention proposes a monostable electromagnetic actuator and an active vibration isolation control method and system thereof, so as to solve the problem of insufficient dynamic response of the existing actuator.
[0005] In a first aspect, the present invention discloses a monostable electromagnetic actuator, comprising a base, a side support seat, a permanent magnetic linear guide rail, a slide rail, an adjustable bolt, a connection mass block, a first elastic member, a coil mover, a second elastic member, and a guide rod; the side support seat is fixedly connected to the base;
[0006] A permanent magnetic linear guide rail is fixedly arranged on the base;
[0007] The slide rail is fixedly arranged on the side support seat and is perpendicular to the permanent magnetic linear guide rail;
[0008] An adjustable bolt is arranged on the top of the side support seat;
[0009] A connecting mass block is slidably arranged on a slide rail;
[0010] A first elastic member, two ends of which are respectively fixedly connected to the bottom of the adjustable bolt and the top of the connection mass block;
[0011] The coil mover has a coil inside and is slidably arranged on a permanent magnetic linear guide rail;
[0012] The second elastic member has one end hinged at the bottom end of the connecting mass block and the other end hinged at the top end of the coil mover; the second elastic member is in the original length state when it is in a vertical state and its axis coincides with the axis of the first elastic member and the vertical axis of the coil mover, and is in a stretched state at other positions.
[0013] The guide rod passes through the through hole on the coil mover and is slidably connected with the coil mover. Both ends of the guide rod are fixedly connected with the base respectively.
[0014] On the basis of the above technical solution, preferably, the side support seat includes a side plate vertically arranged with the base and a connecting plate fixedly connected to the top end of the side plate, the connecting plate is provided with a through hole, and the adjustable bolt passes through the through hole of the connecting plate and is fixedly connected to one end of a sliding connecting block slidingly arranged on the slide rail, and the other end of the sliding connecting block is fixedly connected to one end of the first elastic member.
[0015] On the basis of the above technical solution, preferably, the first elastic member includes a first telescopic rod and a first spring sleeved on the first telescopic rod, two ends of the first spring are respectively fixed to two ends of the first telescopic rod and are elongated or compressed as the first telescopic rod is extended or retracted, and the two ends of the first telescopic rod are respectively fixedly connected to the sliding connecting block and the connecting mass block.
[0016] On the basis of the above technical solution, preferably, the second elastic member includes a second telescopic rod and a second spring sleeved on the second telescopic rod; the two ends of the second spring are respectively fixed to the two ends of the second telescopic rod and are obliquely stretched and restored to their original length as the second telescopic rod is obliquely stretched and extended, and the two ends of the second telescopic rod are respectively rotatably hinged to the connecting mass block and the coil mover.
[0017] On the basis of the above technical solution, preferably, the base includes a bottom plate and two fixing seats fixedly arranged at both ends of the bottom plate, the two fixing seats are symmetrically arranged about the vertical center axis of the bottom plate, and the two ends of the guide rod are respectively fixedly connected to the two fixing seats.
[0018] The second aspect of the present invention discloses an active vibration isolation control method of the monostable electromagnetic actuator according to the first aspect of the present invention, the method comprising:
[0019] S1. Input the given current signal as the control signal into the monostable electromagnetic actuator P(z), use the first neural network to identify the parameters of the actuator, and obtain a reference model reflecting the mapping relationship between the control signal acting on the actuator and the actuator output. ;
[0020] S2. Use the sensor to obtain the reference signal i of the excitation source k And import the reference model , according to the reference model Calculate the reference output r of the actuator k ;
[0021] S3. Obtain the error signal e of the actuator k , based on the actuator reference output r k and the error signal e kAdjusting the parameters of the neural network controller constructed using the second neural network to obtain the neural network controller C(z);
[0022] S4. Copy the neural network controller to obtain a copy controller , the reference signal i k Input the replication controller and calculate the real control signal u k , the control signal u k Input the actuator P(z) and get the actuator output y k ; Set the actuator output y k With the object perturbation n k Subtract and get the error signal e of the actuator k ;
[0023] S5. Repeat steps S2 to S4 to implement vibration reduction control.
[0024] On the basis of the above technical solution, preferably, the step S1 specifically includes:
[0025] The time sampling sequence of the collected control signal constitutes the input layer of the first neural network;
[0026] According to the weight and bias parameters between each node of the hidden layer and the input layer, the hidden layer output is calculated using the ReLU activation function;
[0027] The final output is calculated by the weight and bias parameters between each node in the output layer and the hidden layer;
[0028] Calculate the mean square error between the output and the true response of the actuator;
[0029] Perform error back propagation, calculate the partial derivatives of the mean square error with respect to the weights and biases of each layer, and update the corresponding weights and biases according to the partial derivatives;
[0030] Repeat the above steps until the mean square error reaches the set requirement.
[0031] On the basis of the above technical solution, preferably, in step S3, the error signal e of the actuator k This is the absolute acceleration of the actuator base collected by the sensor.
[0032] On the basis of the above technical solution, preferably, it is characterized in that the step S3 specifically includes:
[0033] Collect the reference output of the reference model k The input layer that constitutes the second neural network;
[0034] According to the weight and bias parameters between the hidden layer and the input layer, the hidden layer output is calculated using the tanh activation function;
[0035] According to the weight and bias parameters between the hidden layer and the output layer, the output control signal is calculated using the sigmoid activation function;
[0036] The actuator error signal e measured by the sensor k Back propagation is performed as the initial error to update the weights and bias parameters of each layer.
[0037] In a third aspect, the present invention discloses a monostable electromagnetic actuator active vibration isolation control system, the system comprising:
[0038] The monostable electromagnetic actuator P(z) described in the first aspect of the present invention;
[0039] Actuator parameter identification module: used to input a given current signal as a control signal into a monostable electromagnetic actuator, use the first neural network to perform parameter identification on the actuator, and obtain a reference model that reflects the mapping relationship between the control signal acting on the actuator and the actuator output ;
[0040] Actuator reference output module: used to obtain the reference signal of the excitation source k Import reference model , according to the reference model Calculate the reference output r of the actuator k ;
[0041] Controller parameter adjustment module: used to obtain the error signal e of the actuator k , based on the actuator reference output r k and the error signal e k adjusting parameters of a neural network controller constructed using a second neural network;
[0042] Actuator control module: used to copy the control parameters of the neural network controller to obtain the copy controller, and convert the reference signal i k Input the replication controller and calculate the real control signal u k , the control signal u k Input the actuator P(z) and get the actuator output y k ;
[0043] A vibration measuring instrument, used to obtain the absolute acceleration of the actuator base as an error signal input into the controller parameter adjustment module;
[0044] The vibration excitation generator is used to generate vibration excitation related signals as reference signals to input into the actuator.
[0045] Compared with the prior art, the present invention has the following beneficial effects:
[0046] 1) The present invention proposes a monostable electromagnetic actuator with an elastic boundary, which is used in active vibration control. In the elastic boundary, an adjustable bolt is used to adjust the position of the connected mass block. The second elastic member maintains its original length at the center line and is in a stretched state at other positions. The coil mover only reaches a stable balance at the midpoint position, that is, a monostable nonlinearity. It has efficient power output performance and can efficiently convert the control input signal into an actuator dynamic response. The monostable electromagnetic actuator can exert excellent vibration isolation performance at high frequencies and can drive its movement at low frequencies by applying an active control current to generate an output force that is opposite to the external excitation, thereby achieving a high-performance vibration reduction effect.
[0047] 2) The present invention designs an adaptive active vibration isolation control method based on a neural network around the above-mentioned monostable electromagnetic actuator with elastic boundaries. The offline identification of the actuator and the real-time dynamic adjustment of the controller parameters are completed by neural network training, the control signal is accurately generated, and the high-efficiency output performance of the actuator is used to achieve high-performance vibration suppression of the target structure. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0049] Figure 1 It is a structural schematic diagram of a monostable electromagnetic actuator of the present invention;
[0050] Figure 2 is a schematic structural diagram of a second elastic member of the present invention;
[0051] Figure 3 It is the principle diagram of the monostable electromagnetic actuator of the present invention;
[0052] Figure 4 This is a flow chart of the active vibration isolation control method of the present invention;
[0053] Figure 5 is a schematic diagram of the network structure of the first neural network of the present invention;
[0054] Figure 6 is a schematic diagram of the network structure of the second neural network of the present invention;
[0055] Figure 7 It is a schematic diagram of the structure of the active vibration isolation control system of the present invention;
[0056] Figure 8 It is a schematic diagram of the active vibration isolation control effect of the present invention.
[0057] In the figure: 1. base; 1-1. bottom plate; 1-2. fixed seat; 2. side support seat; 2-1. side plate; 2-2. connecting plate; 3. permanent magnet linear guide; 4. slide rail; 5. adjustable bolt; 6. sliding connection block; 7. connecting mass block; 8. first elastic member; 8-1. first telescopic rod; 8-2. first spring; 9. coil mover; 10. second elastic member; 10-1. second telescopic rod; 10-2. second spring; 11. guide rod; 12. first rotating shaft; 13. second rotating shaft. DETAILED DESCRIPTION
[0058] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0059] See also Figure 1 The present invention proposes a monostable electromagnetic actuator, including a base 1, a side support seat 2 fixedly connected to the base 1, a permanent magnetic linear guide rail 3 fixedly arranged on the base 1, a slide rail 4 fixedly arranged on the side support seat, an adjustable bolt 5 arranged on the top of the side support seat, a sliding connection block 6 and a connecting mass block 7 slidably arranged on the slide rail, a first elastic member 8 arranged between the sliding connection block 6 and the connecting mass block 7, a coil mover 9 slidably arranged on the permanent magnetic linear guide rail, a second elastic member 10 arranged between the connecting mass block 7 and the coil mover 9, and a guide rod 11 passing through the coil mover 9 and fixed on the base.
[0060] The base 1 includes a bottom plate 1 - 1 and two fixing seats 1 - 2 fixedly arranged at two ends of the bottom plate. The two fixing seats 1 - 2 are symmetrically arranged about the vertical center axis of the bottom plate 1 .
[0061] The side support seat 2 includes a side plate 2-1 vertically arranged with the base 1 and a connecting plate 2-2 fixedly connected with the top end of the side plate 2-1. A through hole is provided on the connecting plate 2-2. The adjustable bolt 5 passes through the through hole of the connecting plate and is fixedly connected with one end of a sliding connecting block 6 slidingly arranged on the slide rail. The other end of the sliding connecting block 6 is fixedly connected with one end of the first elastic member 8.
[0062] The two ends of the first elastic member 8 are respectively fixedly connected to the bottom of the sliding connection block 6 and the top of the connection mass block 7. The first elastic member 8 includes a first telescopic rod 8-1 and a first spring 8-2 sleeved on the first telescopic rod, the two ends of the first spring 8-2 are respectively fixed to the two ends of the first telescopic rod 8-1 and are stretched or compressed as the first telescopic rod 8-1 is extended and contracted, and the two ends of the first telescopic rod 8-1 are respectively fixedly connected to the sliding connection block 6 and the connection mass block 7.
[0063] The second elastic member 10 has one end hinged to the bottom end of the connection mass block 7 through the first rotating shaft 12, and the other end hinged to the top end of the coil mover 9 through the second rotating shaft 13. The structural diagram of the second elastic member 10 is shown in FIG. Figure 2 As shown. The second elastic member 10 includes a second telescopic rod 10-1 and a second spring 10-2 sleeved on the second telescopic rod. The two ends of the second spring 10-2 are respectively fixed to the two ends of the second telescopic rod 10-1 and are stretched and restored to their original length as the second telescopic rod 10-1 is stretched. One end of the second telescopic rod 10-1 is rotatably hinged to the connection mass block 7, and the other end is rotatably hinged to the coil mover 9. When the second elastic member 10 is in a vertical state and its axis coincides with the axis of the first elastic member 8 and the vertical axis of the coil mover 9, the second elastic member 10 is in the original length state, and in a stretched state at other positions, such as Figure 3 shown.
[0064] The coil mover 9 has a coil inside and a through hole in the center. The guide rod 11 passes through the through hole on the coil mover 9 so that the coil mover 9 is slidably set on the guide rod 11. The two ends of the guide rod 11 are respectively fixedly connected to the two fixing seats 1-2 on the base 1.
[0065] like Figure 3 The figure shows the principle diagram of the monostable electromagnetic actuator, wherein the adjustable bolt 5, the sliding connection block 6, the first elastic member 8, and the connection mass block 7 together form an elastic boundary, wherein the adjustable bolt in the elastic boundary is used to adjust the position of the connection mass block 7, and the connection mass block 7 can only move vertically along the slide rail 4 under the action of the first elastic member 8. The slide rail 4 and the permanent magnetic linear guide rail 3 are perpendicular to each other in three-dimensional space.
[0066] The second spring in the second elastic member 10 maintains its original length at the center line and is in a stretched state at other positions, so the coil mover 9 is only at the midpoint ( Figure 3 The spring reaches a stable equilibrium at the center line of the coil, i.e., monostable nonlinearity. At this time, the spring length is the original length d. Among them, the mass of the coil mover 9 is m 1 , the mass of the connected mass block 7 is m 2 , the elastic constant of the first spring is k 1 , the elastic constant of the second spring is k 2The monostable structure has the characteristics of high static stiffness and low dynamic stiffness. A small dynamic force can cause a large amplitude and can stimulate a large amplitude vibration response near the natural frequency. Therefore, the actuator can efficiently convert the control input signal into an output dynamic response. The coil mover 9 moves laterally on the permanent magnetic linear guide 3 and is damped by the damping c 1 When a harmonic current signal i is input into the coil on the coil mover 9, the mover will be subjected to an electromagnetic force F=θi according to the principle of electromagnetic induction, where θ is a constant related to the magnetic flux density and the coil length. The electromagnetic force F drives the coil mover 9 to move, thereby realizing the power output of the actuator.
[0067] The monostable electromagnetic actuator of the present invention can exert excellent vibration isolation performance at high frequencies, and can drive its movement at low frequencies by applying an active control current to generate an output force that is in antiphase with the external excitation, thereby achieving a high-performance vibration reduction effect.
[0068] The present invention performs active vibration isolation control based on the above monostable electromagnetic actuator and proposes an active vibration isolation control method, such as Figure 4 As shown, the method includes:
[0069] S1, input the given current signal as the control signal into the monostable electromagnetic actuator P(z) (i.e. Figure 4 The first neural network is used to identify the parameters of the actuator, and a reference model reflecting the mapping relationship between the control signal acting on the actuator and the actuator output is obtained. .
[0070] like Figure 5 FIG. 1 is a schematic diagram of the network structure of the first neural network, wherein step S1 specifically includes:
[0071] S11, acquisition control signal time sampling sequence The input layer that constitutes the first neural network;
[0072] S12, the hidden layer has 256 nodes in total. According to the weight and bias parameters between each node in the hidden layer and the input layer, the ReLU activation function is used to calculate the hidden layer output;
[0073] S13, the final output is calculated by the weight and bias parameters between the output layer and each node of the hidden layer;
[0074] S14, calculating the mean square error between the output and the actual response of the actuator;
[0075] S15, perform error back propagation, calculate the partial derivatives of the mean square error with respect to the weights and biases of each layer, and update the corresponding weights and biases according to the partial derivatives;
[0076] S16. Repeat the above steps S11 to S15 until the mean square error reaches the set requirement.
[0077] S2. Use the sensor to obtain the reference signal i of the excitation source k And import the reference model , according to the reference model Calculate the reference output r of the actuator k .
[0078] S3. Obtain the error signal e of the actuator k , based on the actuator reference output r k and the error signal e k The parameters of the neural network controller constructed using the second neural network are adjusted to obtain the neural network controller C(z).
[0079] The error signal e of the actuator k This is the absolute acceleration of the actuator base collected by the sensor.
[0080] The network structure of the second neural network is as follows Figure 6 As shown, the step S3 specifically includes:
[0081] S31. Collect reference output r of reference model k The input layer that constitutes the second neural network;
[0082] S32, according to the weight and bias parameters between the hidden layer and the input layer, using the tanh activation function to calculate the hidden layer output;
[0083] S33, according to the weight and bias parameters between the hidden layer and the output layer, using the sigmoid activation function to calculate the output control signal;
[0084] S34, the actuator error signal e measured by the sensor k Back propagation is performed as the initial error to update the weights and bias parameters of each layer.
[0085] S4. Copy the neural network controller to obtain a copy controller , the reference signal i k Input the replication controller and calculate the real control signal u k , the control signal u k Input the actuator P(z) and get the actuator output y k ; Set the actuator output y k With the object perturbation n k Subtract and get the error signal e of the actuator k ;
[0086] S5. Repeat steps S2 to S4 to implement vibration reduction control.
[0087] On the basis of the above monostable electromagnetic actuator and its control method, the present invention also proposes an active vibration isolation control system based on the monostable electromagnetic actuator, such as Figure 7 As shown, the system comprises:
[0088] The present invention Figure 1 The monostable electromagnetic actuator P(z) shown;
[0089] An active vibration isolation control device, wherein the active vibration isolation control device stores an active vibration isolation program developed based on LabVIEW, mainly including an actuator parameter identification module, an actuator reference output module, a controller parameter adjustment module, and an actuator control module;
[0090] The actuator parameter identification module is used to input a given current signal as a control signal into the monostable electromagnetic actuator, use a first neural network to perform parameter identification on the actuator, and obtain a reference model reflecting the mapping relationship between the control signal acting on the actuator and the actuator output. ;
[0091] The actuator reference output module is used to obtain the reference signal i of the excitation source. k Import reference model , according to the reference model Calculate the reference output r of the actuator k ;
[0092] The controller parameter adjustment module is used to obtain the error signal e of the actuator. k , based on the actuator reference output r k and the error signal e k adjusting parameters of a neural network controller constructed using a second neural network;
[0093] The actuator control module is used to copy the control parameters of the neural network controller to obtain a copy controller, and the reference signal i k Input the replication controller and calculate the real control signal u k , the control signal u k Input the actuator P(z) and get the actuator output y k .
[0094] The active vibration isolation control device converts the output y of the actuator k Input the data analysis program developed based on LabVIEW for data analysis and visualization.
[0095] A vibration measuring instrument, used to obtain the absolute acceleration of the actuator base as an error signal input into the controller parameter adjustment module;
[0096] The vibration excitation generator is used to generate vibration excitation related signals as reference signals to input into the actuator.
[0097] When external vibration excitation acts on the controlled structure, the neural network-based adaptive algorithm of the present invention is used to adjust the parameters of the neural network controller, thereby dynamically adjusting the phase and gain of the control current in real time, and inputting the control current to the actuator. The actuator efficiently outputs a force opposite to the vibration interference, directly offsetting the vibration interference force, thereby achieving high-performance vibration suppression of the target structure.
[0098] The above system embodiments are based on the method embodiments. For a brief description of the system embodiments, please refer to the method embodiments.
[0099] Figure 8 FIG. 1 is a control effect diagram of using the monostable actuator of the present invention to perform active vibration isolation control in a certain embodiment. Figure 7 It can be seen that the present invention has a good vibration reduction effect.
[0100] The present invention also discloses an electronic device, comprising: at least one processor, at least one memory, a communication interface and a bus; wherein the processor, memory, and communication interface communicate with each other through the bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to implement the aforementioned method of the present invention.
[0101] The present invention also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, wherein the computer instructions enable the computer to implement all or part of the steps of the method described in the embodiment of the present invention. The storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk, and other media that can store program codes.
[0102] The system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be distributed to multiple network units. A person skilled in the art may select some or all of the modules according to actual needs to achieve the purpose of the solution of this embodiment without creative effort.
[0103] The above description is only a preferred embodiment of the present invention and is 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 in the protection scope of the present invention.
Claims
1. A monostable electromagnetic actuator, comprising a base (1), characterized in that: It also includes a side support seat (2), a permanent magnetic linear guide rail (3), a slide rail (4), an adjustable bolt (5), a connecting mass block (7), a first elastic member (8), a coil mover (9), a second elastic member (10), and a guide rod (11); A side support seat (2) fixedly connected to the base (1); A permanent magnetic linear guide rail (3) is fixedly arranged on the base (1); The slide rail (4) is fixedly arranged on the side support seat (2) and is perpendicular to the permanent magnetic linear guide rail; An adjustable bolt (5) is fixedly arranged on the top of the side support seat (2); A connecting mass block (7) is slidably arranged on the slide rail (4); A first elastic member (8), two ends of which are respectively fixedly connected to the bottom of the adjustable bolt (5) and the top of the connection mass block (7); The coil mover (9) is provided with a coil inside and is slidably arranged on the permanent magnetic linear guide rail (3); A second elastic member (10) has one end hinged to the bottom end of the connection mass block (7) and the other end hinged to the top end of the coil mover (9); The second elastic member (10) is in an original length state when in a vertical state and its axis coincides with the axis of the first elastic member (8) and the vertical axis of the coil mover (9), and is in a stretched state at other positions; A guide rod (11) passes through a through hole on the coil mover (9) and is slidably connected to the coil mover (9), and both ends of the guide rod are respectively fixedly connected to the base (1); The side support seat (2) comprises a side plate (2-1) vertically arranged with the base (1) and a connecting plate (2-2) fixedly connected to the top of the side plate, wherein the connecting plate (2-2) is provided with a through hole, and an adjustable bolt (5) passes through the through hole of the connecting plate and is fixedly connected to one end of a sliding connecting block (6) slidably arranged on the slide rail (4), and the other end of the sliding connecting block (6) is fixedly connected to one end of a first elastic member (8).
2. The monostable electromagnetic actuator according to claim 1, characterized in that: The first elastic member (8) comprises a first telescopic rod (8-1) and a first spring (8-2) sleeved on the first telescopic rod (8-1); two ends of the first spring (8-2) are respectively fixed to two ends of the first telescopic rod (8-1) and are stretched or compressed as the first telescopic rod (8-1) is stretched; and two ends of the first telescopic rod (8-1) are respectively fixedly connected to a sliding connection block (6) and a connection mass block (7).
3. The monostable electromagnetic actuator according to claim 2, characterized in that: The second elastic member (10) comprises a second telescopic rod (10-1) and a second spring (10-2) sleeved on the second telescopic rod (10-1); the two ends of the second spring (10-2) are respectively fixed to the two ends of the second telescopic rod (10-1) and are inclined and extended as the second telescopic rod (10-1) is inclined and extended, and the two ends of the second telescopic rod (10-1) are respectively rotatably hinged to the connection mass block (7) and the coil mover (9).
4. The monostable electromagnetic actuator according to claim 1, characterized in that: The base (1) comprises a bottom plate (1-1) and two fixing seats (1-2) fixedly arranged at two ends of the bottom plate (1-1); the two fixing seats (1-2) are symmetrically arranged about the vertical center axis of the bottom plate (1-1); and the two ends of the guide rod (11) are respectively fixedly connected to the two fixing seats (1-2).
5. An active vibration isolation control method for a monostable electromagnetic actuator according to any one of claims 1 to 4, characterized in that: The method comprises: S1. Input the given current signal as the control signal into the monostable electromagnetic actuator P(z), use the first neural network to identify the parameters of the actuator, and obtain a reference model reflecting the mapping relationship between the control signal acting on the actuator and the actuator output. ; S2. Use the sensor to obtain the reference signal i of the excitation source k And import the reference model , according to the reference model Calculate the reference output r of the actuator k ; S3. Obtain the error signal e of the actuator k , based on the actuator reference output r k and the error signal e k Adjust the parameters of the neural network controller constructed using the second neural network to obtain the neural network controller ; S4. Copy the neural network controller to obtain a copy controller , the reference signal i k Input the replication controller and calculate the real control signal u k , the control signal u k Input the actuator P(z) and get the actuator output y k ; Set the actuator output y k With the object perturbation n k Subtract and get the error signal e of the actuator k ; S5. Repeat steps S2 to S4 to implement vibration reduction control.
6. The active vibration isolation control method of a monostable electromagnetic actuator according to claim 5, characterized in that: The step S1 specifically includes: The time sampling sequence of the collected control signal constitutes the input layer of the first neural network; According to the weight and bias parameters between each node of the hidden layer and the input layer, the hidden layer output is calculated using the ReLU activation function; The final output is calculated by the weight and bias parameters between each node in the output layer and the hidden layer; Calculate the mean square error between the output and the true response of the actuator; Perform error back propagation, calculate the partial derivatives of the mean square error with respect to the weights and biases of each layer, and update the corresponding weights and biases according to the partial derivatives; Repeat the above steps until the mean square error reaches the set requirement.
7. The active vibration isolation control method of a monostable electromagnetic actuator according to claim 5, characterized in that: In step S3, the error signal e of the actuator k This is the absolute acceleration of the actuator base collected by the sensor.
8. The active vibration isolation control method of a monostable electromagnetic actuator according to claim 7, characterized in that: The step S3 specifically includes: Collect the reference output of the reference model k The input layer that constitutes the second neural network; According to the weight and bias parameters between the hidden layer and the input layer, the hidden layer output is calculated using the tanh activation function; According to the weight and bias parameters between the hidden layer and the output layer, the output control signal is calculated using the sigmoid activation function; The actuator error signal e measured by the sensor k Back propagation is performed as the initial error to update the weights and bias parameters of each layer.
9. An active vibration isolation control system based on a monostable electromagnetic actuator, characterized in that: The system comprises: The monostable electromagnetic actuator P(z) according to any one of claims 1 to 4; Actuator parameter identification module: used to input a given current signal as a control signal into a monostable electromagnetic actuator, use the first neural network to perform parameter identification on the actuator, and obtain a reference model that reflects the mapping relationship between the control signal acting on the actuator and the actuator output ; Actuator reference output module: used to obtain the reference signal of the excitation source k Import reference model , according to the reference model Calculate the reference output r of the actuator k ; Controller parameter adjustment module: used to obtain the error signal e of the actuator k , based on the actuator reference output r k and the error signal e k adjusting parameters of a neural network controller constructed using a second neural network; Actuator control module: used to copy the control parameters of the neural network controller to obtain the copy controller, and convert the reference signal i k Input the replication controller and calculate the real control signal u k , the control signal u k Input the actuator P(z) and get the actuator output y k ; Vibration measuring instrument: used to obtain the absolute acceleration of the actuator base as an error signal input into the controller parameter adjustment module; Vibration excitation generator: used to generate vibration excitation related signals as reference signals to input into the actuator.
