CSRR-Based Microwave Two-Dimensional Displacement Sensor and Its Numerical Simulation Optimization Method

Abstract

The present invention discloses a microwave two-dimensional displacement sensor based on CSRR and its numerical simulation optimization method. The sensor of the present invention includes a stator and a rotor. The stator includes: a dielectric plate in the middle layer, a microstrip made of metal is provided on the top layer of the dielectric plate, a metal thin sheet is provided on the bottom layer of the dielectric plate, and two rectangular slits are engraved on the metal thin sheet; there are two rotors in total, both of which are composed of two parts: the lower part is a triangular metal patch, and the upper part uses a triangular dielectric plate; the metal patch is arranged on the lower surface of the upper layer; the metal patches of the two rotors are respectively in electrical contact with the two rectangular slit units of the stator and the rotor can move along the bottom surface of the stator, jointly constituting a CSRR whose electrical properties change according to the movement of the stator. The sensor structure proposed by the present invention is more compact and makes greater use of the space on the dielectric plate.

Description

Technical Field

[0001] The present invention belongs to the technical field of microwave sensors, and particularly relates to a microwave two-dimensional displacement sensor based on CSRR and a numerical simulation optimization method thereof. Background Art

[0002] Microwave sensors have the advantages of high sensitivity, stability, low cost, etc., and play an important role in many fields such as medical treatment, biomedicine, and industry. Based on different principles, researchers have developed microwave sensors with functions such as material identification, humidity sensing, and material defect detection.

[0003] Since the measurement of the relative displacement of the target is very important in many applications, such as the field of aerospace vehicles, etc., in recent years, displacement sensors based on various principles with different forms have emerged to complete this task. The general strategy of a microwave displacement sensor is that the object to be measured is connected to a movable structure of a part of the sensor, which may be a resonant structure or a part of a microwave circuit. The movement of the object to be measured will carry the movable part of the sensor together, and the movement of this part will change the properties of a part of the microwave circuit or its resonant unit or trigger different coupling effects. Furthermore, information about the movement amount of the object to be measured can be extracted from the output signal obtained at the sensor port.

[0004] Microwave displacement sensors require higher sensitivity and smaller size. At the same time, most of the previous microwave displacement sensors can only measure the relative displacement of the target object in one dimension, and the designed size is quite large relative to its dynamic range. Therefore, it is an urgent problem to develop a more compact microwave sensor that can measure the displacement of the target object in a two-dimensional plane.

[0005] In addition, for microwave sensors containing multiple parameters, parameter tuning is usually a laborious and time-consuming mechanical work if carried out manually. Summary of the Invention

[0006] In view of the above deficiencies in the prior art, the present invention provides a two-dimensional microwave displacement sensor based on CSRR (Complementary Split Ring Resonator) and a numerical simulation optimization method thereof. The present invention can provide a relatively large measurement range on a two-dimensional plane with a smaller device size. For some parameters that need to be optimized, the present invention uses the particle swarm algorithm to optimize these parameters.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] Microwave two-dimensional displacement sensor based on CSRR, which includes a stator and a rotor; the stator includes: the middle layer is a dielectric plate, the top layer of the dielectric plate is provided with a microstrip made of metal, the bottom layer of the dielectric plate is provided with a metal sheet, and two rectangular slits are engraved on the metal sheet; there are two rotors in total, both of which are composed of two parts: the lower part is a triangular metal patch, and the upper part is a triangular dielectric plate; the metal patch is arranged on the lower surface of the upper layer; the metal patches of the two rotors are respectively in electrical contact with the two rectangular slit units of the stator and the rotor can move along the bottom surface of the stator, jointly forming a CSRR whose electrical properties change according to the movement of the stator.

[0009] Preferably, the microstrip is in the shape of a right-angled folded edge, the two right-angled sides are respectively rectangular, the two right-angled sides are respectively parallel to two adjacent sides of the stator, and the outer sides at the transition of the two right-angled sides are chamfered.

[0010] Preferably, the two ends of the microstrip are respectively close to the edge ports of the dielectric plate, which are the input port and the output port respectively, forming a two-port network. The input port and the output port are respectively connected to an SMA head.

[0011] Preferably, the bottom layer of the stator uses a metal sheet containing two rectangular slit units of different sizes as the bottom surface of the entire displacement sensor.

[0012] Preferably, the dielectric constant of the stator dielectric plate is 3.66, the loss tangent is 0.004, and the thickness is 0.762 mm.

[0013] Preferably, the length of the stator dielectric plate is 20 mm and the width is 20 mm.

[0014] Preferably, the material of the rotor dielectric plate is the same as that of the stator dielectric plate.

[0015] In addition to the above-mentioned parameters, the microwave two-dimensional displacement sensor proposed by the present invention also has some other parameters: the parameters of the inner circle shape of the rectangular slit in the bottom metal sheet: slit width, position of the circle, size of the circle. These three parameters of the two rectangular slits are different from each other, thus forming a parameter combination to be optimized with a total of six parameters. During the displacement process of the target object on the two-dimensional plane, the electrical properties of the CSRR will be changed. The two CSRR units respectively measure the displacements in the x direction and the y direction on the two-dimensional plane, and the displacement in the y direction will also affect the measurement in the x direction. The optimization goal is to minimize the measurement error caused by this influence. Through the joint simulation optimization method, the above work is completed. The specific steps are:

[0016] S1. Set the parameter combination to be optimized for the microwave displacement sensor, that is, the total of six parameters described above;

[0017] S2. Initialize the program parameter values;

[0018] S3. Construct a set of sensor individuals based on a series of feasible parameter value combinations generated in the design space by the particle swarm algorithm to form a population;

[0019] S4. Build a model and conduct simulations, and calculate the fitness value of the individual displacement sensor based on the data obtained from the simulations;

[0020] S5. Update the individual optimal parameter combination, the population optimal parameter combination, and the velocity matrix in the particle swarm algorithm during the iteration process according to the fitness value;

[0021] S6. Determine whether the number of iterations has reached the maximum number of iterations; if not, repeat the above steps S3 - S5, if so, output the optimized sensor parameter values.

[0022] Preferably, in step S2, the program parameter values include the weights of several quantities such as the individual best position, the population best position, and inertia considered during particle update; the maximum number of iterations; the geometric factors of the sensor design and the constraints of the manufacturing process on the design space.

[0023] Preferably, in step S3, the specific process of generating feasible parameter value combinations is that during the particle swarm optimization process, the next position of the particle is updated according to the several weights specified in step S2; the boundary of the design space is specified by step S2, and for particles that exceed the boundary range, they are attached to the boundary.

[0024] Preferably, in step S4, for the fitness of the population individuals where Cost x = f x|y＝0 - f x|y＝2 , that is, the magnitude of the error caused by the displacement in the y - direction to the resonant frequency measured in the x - direction. The smaller such an error, the higher the fitness. Similarly, Cost y = f y|x＝0 - f y|x＝2 . Where F fitness refers to the fitness of the individual; the fitness of the individual is obtained by taking the reciprocal of the sum of the errors Cost x , Cost y in the x and y directions. The error Cost x in the x - direction is calculated from the difference between the resonant frequencies when the displacement in the x - direction is fixed and the displacement in the y - direction is 0, with the resonant frequency being f x|y＝0 , and when the displacement in the y - direction is 2 mm, with the resonant frequency being f x|y＝2 . Similarly, the error Cost y in the y - direction is calculated from the difference between the resonant frequencies when the displacement in the y - direction is fixed and the displacement in the x - direction is 0, with the resonant frequency being f y|x＝0 , and when the displacement in the x - direction is 2 mm, with the resonant frequency being f y|x＝2 .

[0025] Preferably, in step S5, the update steps of the individual optimal parameter combination are as follows: compare the historical fitness optimum of all individuals with the fitness of this iteration. If the fitness of this iteration is better, update the individual optimal parameter combination. The update steps of the population optimal parameter include: compare the individual optimal fitness values of all individuals with the population optimal fitness value during the iteration process of the population. If there is an individual optimal fitness value that is better than the population optimal fitness value, record it as the population optimal fitness value, and record the parameter combination corresponding to this individual optimal fitness value.

[0026] Preferably, the velocity update in step S5 is represented by the following formula:

[0027]

[0028]

[0029] w is the inertia factor, which is used to control the movement range of the particle and is related to the convergence speed; p i , p g are the individual best position and the population best position respectively; c 1 , c 2 control the influence of the particle by the individual optimal position and the population optimal position respectively; r 1 , r 2 is a random number on [0, 1], and α controls the weight of the velocity.

[0030] Compared with the prior art, the present invention has the following advantages:

[0031] First, the sensor structure proposed by the present invention is more compact, making greater use of the space on the dielectric plate.

[0032] Second, the present invention can sense displacements in two dimensions. Compared with a one-dimensional sensor that can only detect the displacement of a target object on a straight line, the sensor proposed by the present invention can sense the displacement of a target object on a plane.

[0033] Third, the present invention uses the particle swarm algorithm to optimize the sensor parameter combination. Compared with mechanical manual parameter adjustment, it is more efficient and more likely to find the global optimal value. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a schematic top view of the stator of the present invention;

[0035] Figure 2 is a schematic bottom view of the stator of the present invention;

[0036] Figure 3 is a side view of the stator of the present invention;

[0037] Figure 4 is a schematic structural diagram of the mover of the present invention;

[0038] Figure 5 is a flowchart of the co-simulation optimization method for sensor parameter values;

[0039] Figure 6 is a diagram for marking the optimized parameters of the present invention;

[0040] Figure 7 is a schematic diagram showing the change of the two-port transmission coefficient of the present invention with the displacement of the mover in the x direction;

[0041] Figure 8 is a schematic diagram showing the change of the two-port transmission coefficient of the present invention with the displacement of the mover in the y direction. Specific implementation method

[0042] The present invention will be further described in detail below with reference to the specific embodiments in the accompanying drawings.

[0043] As Figure 1-4 shown is a schematic structural diagram of a preferred embodiment of the present invention. The microwave displacement sensor with high dynamic range in this embodiment includes a stator and a mover, and their structures are described as follows.

[0044] The stator includes: the intermediate layer 1-2 is made of a dielectric board of Rogers 4350 series; a microstrip 1-1 made of metal is provided on the top layer of the dielectric board; a metal thin sheet 1-3 is provided on the bottom layer of the dielectric board, and two rectangular slit units 1-4, 1-5 with different sizes are engraved on the metal thin sheet. The inner circle of the slit is an irregular figure.

[0045] The microstrip on the top layer of the stator is in the shape of a right-angled folded edge. The two right-angled sides are respectively rectangular, and the two right-angled sides are respectively parallel to two adjacent sides of the stator. The outer sides at the transition of the two right-angled sides are chamfered. The two ends of the microstrip close to the edge of the dielectric board 1-2 are respectively the input port 1-6 and the output port 1-7, forming a two-port network. The two ports are respectively connected to an SMA head, and the SMA head is connected to a vector network analyzer. The above microstrip structure parameters are all set according to the standard of 50 ohms to match the external measurement circuit and prevent losses.

[0046] The metal thin sheet 1-3 on the bottom layer is arranged on the lower surface of the intermediate dielectric board 1-2, and two rectangular slits 1-4, 1-5 that will cooperate with the mover to measure the displacement are engraved on the metal thin sheet.

[0047] The mover is connected to the object to be measured to measure its displacement relative to the stator. There are two movers in total, both connected to the object to be measured; the two movers are of different sizes but have generally the same structure and are both composed of two parts: the lower part is a triangular metal patch 2-1, 3-1; the upper part uses triangular dielectric plates 2-2, 3-2; the metal patches are arranged on the lower surface of the upper layer. The metal patches 2-1, 3-1 of the two movers are in direct contact with a rectangular slit unit 1-4, 1-5 of the stator respectively, and together form a CSRR whose electrical properties change according to the movement of the stator.

[0048] The two ends of the mover metal patches 2-1, 3-1 are in contact with the two rectangular slits 1-4, 1-5 of the stator respectively to form an electrical contact. The mover can move in any direction on the bottom surface of the stator. By adjusting the relative displacement of the mover along the stator to the CSRR jointly formed by the rectangular slit of the stator and the triangular metal patch of the mover, the frequency point where the transmission zero point of the output is located is affected.

[0049] In this embodiment, the stator dielectric layer 1-2 has a dielectric constant of 3.66, a loss tangent of 0.004, and a thickness of 0.762 mm. The entire length of the stator dielectric plate is 20 mm, and the width is 20 mm. The dielectric plate material of the mover is the same as that of the stator.

[0050] As Figure 5 shown, this embodiment provides a method for jointly simulating and optimizing the numerical value of a microwave sensor, including the following steps:

[0051] S1. Set the parameter combination to be optimized for the microwave displacement sensor, that is, the six parameters mentioned above, namely: the parameters of the inner shape of the rectangular slit in the bottom metal sheet 1-3: slit width, position of the circle, size of the circle. These three parameters of the two rectangular slits are different from each other, thus constituting a parameter combination to be optimized with a total of six parameters;

[0052] S2. Initialize the program parameter values;

[0053] S3. According to a series of feasible parameter value combinations generated by the particle swarm algorithm in the design space, construct a set of sensor individuals to form a population;

[0054] S4. Build a model and perform simulation, and calculate the fitness value of the individual displacement sensor according to the data obtained from the simulation;

[0055] S5. Update the individual optimal parameter combination, the population optimal parameter combination, and the velocity matrix in the particle swarm algorithm during the iteration process according to the fitness value;

[0056] S6. Determine whether the number of iterations has reached the maximum number of iterations; if not, repeat the above steps S3-S5, if so, output the optimized sensor parameter combination.

[0057] Refer to Figure 6 , there are a total of six parameters to be optimized.

[0058] In step S2, the program parameter values include the maximum number of iterations, the inertia factor, and the weights of the influence of the particle on the individual optimal and the global optimal positions.

[0059] In step S3, the initial individual set randomly generates a feasible parameter combination in the design space by a numerical calculation software, and a sensor is constructed from this combination; in subsequent iterations, each round of the individual set is given by the rules of position update and velocity update of the particle swarm; for the parameter combinations that exceed the feasible design space range during the particle optimization process, they are placed at the boundary of the design space.

[0060] Step S4 specifically includes: generating different sensor model scripts, the full-wave electromagnetic simulation software reads the script file to automatically construct the model and simulate, and the obtained simulation data is used to calculate the fitness value by the fitness function. The formula for the individual fitness evaluation function in the population is as follows:

[0061]

[0062] Cost x = f x|y＝0 - f x|y＝2

[0063] Cost y = f y|x＝0 - f y|x＝2

[0064] The fitness of an individual is obtained by taking the reciprocal of the sum of the errors in the x and y directions.

[0065] In step S5, the steps for updating the individual optimum include: comparing the current fitness values of all individuals with their respective optimum fitness values during the iteration process. If the current fitness value is better, record the current fitness value as the individual optimum fitness value and record the current individual design parameters, which are the individual optimum design parameters; the steps for updating the global optimum include: comparing all the individual optimum fitness values with the global optimum fitness value of the population during the iteration process. If there is an individual optimum fitness value that is better than the global optimum fitness value, record it as the global optimum fitness value and record the design parameters corresponding to the individual optimum fitness value, which are the global optimum design parameters.

[0066] As Figure 6 shown in the parameter annotation diagram, a total of six parameters marked as shown are optimized during the optimization process.

[0067] As Figure 7The S-parameter diagram of the sensor under the displacement in the x direction is shown. It can be seen that the error in the y direction is very small, and the resonance point on the left changes with the displacement in the x direction.

[0068] As Figure 8 The S-parameter diagram of the sensor under the displacement in the y direction is shown. It can be seen that the error in the x direction is very small, and the resonance point on the right changes with the displacement in the y direction.

[0069] Through the swarm intelligence optimization algorithm, the present invention optimizes the parameter combination in the design space of the sensor in a suitable process. Compared with the aimless global search, it has higher efficiency, and the result is more likely to be the global optimal solution. The particle swarm algorithm is a numerical optimization algorithm. Compared with some other swarm intelligence algorithms, it has many characteristics suitable for the optimization of sensor parameters, such as fewer hyperparameters, faster convergence speed, and a simple and clear process that is easy to understand. The particle swarm algorithm can be combined with the finite element simulation to optimize the sensor parameters in its design space.

[0070] The above is only the preferred embodiment of the present invention and the applied technical principle. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described here. Various obvious changes, re-adjustments, and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, it can also include more other equivalent embodiments, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A microwave two-dimensional displacement sensor based on CSRR, comprising a stator and a rotor, Characterized in that: The stator includes: the middle layer is a dielectric plate, the top layer of the dielectric plate is provided with a microstrip made of metal, the bottom layer of the dielectric plate is provided with a metal sheet, and two rectangular slits are engraved on the metal sheet; There are two rotors in total, both of which are composed of two parts: the lower part is a triangular metal patch, and the upper part is a triangular dielectric plate. The triangular metal patch is arranged on the lower surface of the upper layer; the two triangular metal patches are respectively in electrical contact with the two rectangular slits of the metal sheet and the rotor can move along the bottom surface of the stator, jointly constituting a CSRR that changes its electrical properties according to the movement of the stator.

2. The microwave two-dimensional displacement sensor based on CSRR according to claim 1, Characterized in that: The microstrip is in the shape of a right-angled folded edge, the two right-angled sides are respectively rectangular, the two right-angled sides are respectively parallel to two adjacent sides of the stator, and the outer sides at the transition of the two right-angled sides are chamfered.

3. The microwave two-dimensional displacement sensor based on CSRR according to claim 1, Characterized in that: The two ends of the microstrip close to the edge of the dielectric plate are respectively an input port and an output port, forming a two-port network.

4. The microwave two-dimensional displacement sensor based on CSRR according to claim 1, Characterized in that: The dielectric constant of the stator dielectric plate is 3.66, the tangent of the loss angle is 0.004, and the thickness is 0.762 mm.

5. The microwave two-dimensional displacement sensor based on CSRR according to claim 1, Characterized in that: The length of the stator dielectric plate is 20 mm and the width is 20 mm.

6. A numerical simulation optimization method for the microwave two-dimensional displacement sensor according to any one of claims 1-5, Characterized in that According to the following steps: S1. Set the parameter combination to be optimized for the microwave two-dimensional displacement sensor. The parameters are the parameters related to the inner shape of the rectangular slit in the metal sheet: slit width, position of the circle, size of the circle; these three parameters of the two rectangular slits are different from each other, thus constituting a parameter combination to be optimized with a total of six parameters; S2. Initialize the program parameter values; S3. According to a series of parameter value combinations generated by the particle swarm algorithm in the design space, construct a set of sensor individuals to form a population; S4. Build a model and perform simulation, and calculate the fitness value of the individual displacement sensor according to the data obtained from the simulation; S5. Update the individual optimal parameter combination, the population optimal parameter combination and the velocity matrix in the particle swarm algorithm during the iteration according to the fitness value; S6. Judge whether the number of iterations reaches the maximum number of iterations; if not, repeat the above steps S3-S5, if so, output the optimized sensor parameter combination.

7. The numerical simulation optimization method according to claim 6, Characterized in that: In step S2, the program parameter values include the maximum number of iterations, the inertia factor, and the weight of the influence of the particle on the individual optimal and population optimal positions.

8. The numerical simulation optimization method according to claim 6, Characterized in that: In step S3, the initial individual set randomly generates parameter combinations in the design space by a numerical calculation software, and a sensor is constructed from this combination; in subsequent iterations, each round of the individual set is given by the rules of position update and velocity update of the particle swarm; for parameter combinations that exceed the design space range during the particle optimization process, they are placed at the boundary of the design space.

9. The numerical simulation optimization method according to claim 6, characterized in that: step S4 specifically includes: generating different sensor model scripts, the full-wave electromagnetic simulation software reads the script file to automatically construct a model and simulate, and the obtained simulation data is used to calculate the fitness value by the fitness function; the formula of the individual fitness evaluation function in the population is as follows: Cost x = f x|y＝0 - f x|y＝2 Cost y = f y|x＝0 - f y|x＝2 Among them, F fitness refers to the fitness of an individual; the fitness of an individual is obtained by taking the reciprocal of the sum of the errors Cost x and Cost y in the x and y directions. The error Cost x in the x direction is calculated from the difference between the resonant frequencies when the displacement in the x direction is fixed and the displacements in the y direction are 0 with a resonant frequency of f x|y＝0 and when the displacement in the y direction is 2 mm with a resonant frequency of f x|y＝2 . Similarly, the error Cost y in the y direction is calculated from the difference between the resonant frequencies when the displacement in the y direction is fixed and the displacements in the x direction are 0 with a resonant frequency of f y|x＝0 and when the displacement in the x direction is 2 mm with a resonant frequency of f y|x＝2 .

10. The numerical simulation optimization method according to claim 6, characterized in that: In step S5, the steps of individual best update include: comparing the current fitness values of all individuals with their respective best fitness values during the iteration process. If the current fitness value is better, record the current fitness value as the individual best fitness value and record the current individual design parameters, which are the individual best design parameters; the steps of global best update include: comparing all individual best fitness values with the global best fitness value of the population during the iteration process. If there is an individual best fitness value that is better than the global best fitness value, record it as the global best fitness value and record the design parameters corresponding to the individual best fitness value, which are the global best design parameters.

Images