Full-process numerical simulation method and system for Hall thruster

Abstract

The invention relates to the field of computer aided design, in particular to a full-flow numerical simulation method and system for a Hall thruster, and the method comprises the steps: obtaining a target region in a current simulation process, predicting a first prediction value of each grid at a next moment through a time sequence of a simulation value of each grid in the target region, recording the simulation value of each grid in the same target area at the next moment in the previous simulation process as a second prediction value of each grid in the current simulation process; and in the current simulation process, obtaining an evolution region, and carrying out weighted fusion on the first predicted values and the second predicted values of all the grids in the target region to obtain simulation values of all the grids in the target region at the next moment. The accuracy of the simulation process is ensured while the simulation time is shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of computer-aided design, and particularly to a full-process numerical simulation method and system for Hall thrusters. Background Art

[0002] Using the method of computer-aided design to simulate the working process of Hall thrusters is a common method for designing efficient Hall thrusters. For example, the particle-in-cell (PIC) method is used to simulate the three-dimensional model of Hall thrusters, and the optimal Hall thruster parameters (such as the structure of the discharge chamber, the distribution of electromagnetic fields, etc.) with the best indicators (such as thrust, specific impulse, working efficiency, service life, etc.) are obtained by simulating the working process of Hall thrusters under multiple sets of parameters.

[0003] In order to accurately simulate the working process of Hall thrusters and obtain the Hall thruster parameters under the optimal indicators, it is often necessary to divide the three-dimensional model of Hall thrusters into a large number of grids and simulate them successively under multiple sets of Hall thruster parameters. However, this will result in huge computational amounts and consume a large amount of simulation time. Although the simulation time can be shortened by reducing the number of grids and parameters of the three-dimensional model, this will lead to inaccurate simulation processes and the inability to obtain the Hall thruster parameters under the globally optimal indicators. Summary of the Invention

[0004] To solve the problem that the simulation time and accuracy cannot be both achieved, the present invention provides a full-process numerical simulation method and system for Hall thrusters.

[0005] The full-process numerical simulation method and system for Hall thrusters of the present invention adopt the following technical solutions: An embodiment of the present invention provides a full-process numerical simulation method for Hall thrusters, and the method includes the following steps: Changing the parameters of the Hall thruster multiple times, and performing a simulation each time after the parameters are changed; when performing each simulation, a simulation value is generated for each grid in the three-dimensional model of the Hall thruster at each moment; During the current simulation process, obtain the target area, where the simulation value differences of all grids in the target area are the smallest, and the average simulation value of all grids in the target area has the same change trend as the average simulation value of all grids in the same target area during the previous simulation process; use the time series of the simulation values of each grid in the target area to predict the first predicted value of each grid at the next moment, and record the simulation value of each grid at the next moment in the same target area during the previous simulation process as the second predicted value of each grid during the current simulation process; During the current simulation process, obtain the evolution area, where the evolution area is adjacent to the target area and has the largest difference from the average simulation value of the target area; The first predicted values and the second predicted values of all grids in the target area are weighted and fused to obtain the simulated values of all grids in the target area at the next moment. The direction of the maximum distribution difference of the simulated values of all grids in the target area at the next moment is consistent with the direction of the maximum distribution difference of the simulated values of all grids in the evolution area.

[0006] Preferably, for obtaining the target area, the difference in the simulated values of all grids in the target area is the smallest, and the average simulated value of all grids in the target area has the same change trend as the average simulated value of all grids in the same target area during the previous simulation process. The specific steps included are as follows: At the current moment in the current simulation process, perform K-Means clustering on the feature points of all grids at the current moment to obtain several first categories; the area formed by all grids in each first category is denoted as a first area; in the current simulation process, obtain the first average value of the simulated values of all grids in the first area at each moment; in the current simulation process, the first average values at all moments before the current moment form the first sequence of each first area; During the previous simulation process, obtain the second average value of the simulated values of all grids in each first area. During the previous simulation process, the second average values at all moments before the current moment form the second sequence of each first area; Obtain the trend similarity between the first sequence and the second sequence of each first area. When the trend similarity is greater than or equal to the first preset threshold, it is determined that: the average simulated value of all grids in the first area has the same change trend as the average simulated value of all grids in the same area during the previous simulation process, and the first area is denoted as the target area.

[0007] Preferably, for obtaining the evolution area, the evolution area is adjacent to the target area and has the largest difference from the average simulated value of the target area. The specific steps included are as follows: During the current simulation process, among all the first areas, obtain all the first areas adjacent to each target area, denoted as target adjacent areas; obtain the average simulated value of all grids in each target adjacent area; Obtain the difference between the average simulated value of the target area and the average simulated value of the target adjacent area, and denote the several target adjacent areas with the largest differences as the evolution areas of each target area.

[0008] Preferably, for weighting and fusing the first predicted values and the second predicted values of all grids in the target area to obtain the simulated values of all grids in the target area at the next moment, the specific steps included are as follows: Preset several groups of fusion weights. For each group of fusion weights (w1, w2), use w1 and w2 to respectively perform weighted summation on the first prediction values and the second prediction values of all grids within the target area. w1 is the weighted value of the first prediction values of all grids, and w2 is the weighted value of the second prediction values of all grids, to obtain the comprehensive prediction simulation values of all grids. Obtain the maximum distribution difference direction of the comprehensive prediction simulation values of all grids within the target area, briefly denoted as the first direction; the maximum distribution difference direction of the simulation values of all grids within the evolution area is briefly denoted as the second direction; obtain the selection index of each group of fusion weights by using the similarity between the first direction and the second direction; obtain the group of fusion weights with the largest selection index, and use this group of fusion weights to perform weighted summation on the first prediction values and the second prediction values of all grids. The obtained comprehensive prediction simulation values of all grids are used as the simulation values of all grids within the target area at the next moment during the current simulation process.

[0009] Preferably, the specific steps for obtaining the maximum distribution difference direction of the simulation values of all grids within the evolution area are as follows: Use the K-Means algorithm to cluster the feature points of all grids within the evolution area to obtain several second categories, obtain the average simulation value within each second category, obtain the difference between the average simulation values of any two second categories, obtain several pairs of second categories with the largest difference, and the direction where the displacement vector is located between the category center points of several pairs of second categories is used as the maximum distribution difference direction of the simulation values of all grids within the evolution area.

[0010] Preferably, the specific steps for obtaining the trend similarity between the first sequence and the second sequence of each first area are as follows: Perform linear normalization processing on the first sequence and the second sequence respectively. Denote the DTW distance between the linearly normalized first sequence and the second sequence as x, and denote exp(-x) as the trend similarity. exp() represents the exponential function with the natural constant as the base.

[0011] Preferably, the feature point of the grid is a vector formed by splicing the center point coordinates of the grid and the simulation value of the grid.

[0012] Preferably, the specific steps for obtaining the selection index of each group of fusion weights by using the similarity between the first direction and the second direction are as follows: Perform KM matching on all the first directions and all the second directions to obtain several matching pairs. Each matching pair contains a first direction and a second direction, and calculate the cosine similarity between the first direction and the second direction of each matching pair; the sum of the cosine similarities of all matching pairs is denoted as the selection index of each group of fusion weights.

[0013] Preferably, the parameters include the inner diameter and outer diameter of the annular discharge chamber of the Hall thruster at different positions, the current magnitude of the coil, the voltage between the acceleration electrodes, the propellant flow density of the particle emitter, and the power of the Hall thruster.

[0014] Another embodiment of the present invention provides a full-process numerical simulation system for a Hall thruster, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the full-process numerical simulation method for a Hall thruster as described above.

[0015] The beneficial effects of the technical solution of the present invention are as follows: The present invention performs weighted fusion on the first prediction values and the second prediction values of all grids in the target area to obtain the simulation values of all grids in the target area at the next moment. This process fuses the simulation results of the previous simulation process and the current simulation process, and directly obtains the simulation values at the next moment of the current simulation process, avoiding the problems of large computational amount and long computational time when simulating a large number of grids by the particle-in-cell (PIC) method. On the other hand, it enables accurate simulation results under multiple sets of Hall thruster parameters to be obtained in a relatively short time (the more grids, the more accurate).

[0016] Among them, the first prediction value is predicted using the time series of the simulation values of each grid in the target area. The simulation values of all grids in the target area have the smallest difference, and the average simulation value of all grids in the target area has the same change trend as the average simulation value of all grids in the same target area during the previous simulation process. This process takes into account that since the differences between the simulation values of all grids in the target area are relatively small, the processes of particle collision, ionization, acceleration, etc. of the grids in the target area may not be interfered by the processes of particle collision, ionization, acceleration, etc. of the particles in other grids. Therefore, during the current simulation process, the time series of the simulation values of each grid in the target area is used to predict the first prediction value of each grid at the next moment, which provides an accurate reference value for the simulation value of each grid at the next moment.

[0017] Among them, since the evolution area is adjacent to the target area and has the largest average simulation value, there is an obvious correlation between the target area and the evolution area. For example, there are situations such as particle exchange, charge offset, and evolution of particle velocity distribution due to particle collision, ionization, and acceleration between the target area and the evolution area.

[0018] Based on the evolution region, further make the direction of the maximum distribution difference of the simulated values of all grids in the target region at the next moment consistent with the direction of the maximum distribution difference of the simulated values of all grids in the evolution region. For this process, from a larger range (i.e., the larger local region composed of the target region and the evolution region), when there is a direction of change spread such as the velocity, charge, ionization situation, etc. of the particles, it will not mutate in a short time. By making the predicted direction of change spread consistent with the direction of change spread in the evolution region, the consistency of the direction of change spread in a larger range (i.e., no mutation) is ensured, and thus the accuracy of the predicted simulated values is ensured.

[0019] In summary, the present invention reduces the simulation time while ensuring the accuracy of the simulation process, enabling the efficient acquisition of Hall thruster parameters that meet the design expectations using the method of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0021] Figure 1 It is a flowchart of the steps of the full-process numerical simulation method for a Hall thruster provided by an embodiment of the present invention; Figure 2 It is the first design scheme of the discharge chamber of the Hall thruster provided by an embodiment of the present invention; Figure 3 It is the second design scheme of the discharge chamber of the Hall thruster provided by an embodiment of the present invention; Figure 4 It is the third design scheme of the discharge chamber of the Hall thruster provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, in conjunction with the accompanying drawings and preferred embodiments, describe in detail the specific implementation manners, structures, features, and effects of the full-process numerical simulation method and system for a Hall thruster proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs.

[0024] The following specifically describes the specific solutions of the full - process numerical simulation method and system for Hall thrusters provided by the present invention in conjunction with the accompanying drawings.

[0025] Example 1: Please refer to Figure 1 , which shows the flowchart of the steps of the full - process numerical simulation method for Hall thrusters provided by an embodiment of the present invention. The method includes the following steps: Step S101: Perform full - process numerical simulation using a three - dimensional model of a Hall thruster. During the simulation, a simulation value is generated for each grid in the three - dimensional model of the Hall thruster at each moment.

[0026] The Hall thruster uses electromagnetic fields to ionize and accelerate neutral particles (such as xenon atoms) to form thrust. The settings of parameters such as the structure of the Hall thruster (e.g., the inner and outer diameters at different positions of the annular discharge chamber), the distribution of the electromagnetic field (e.g., the position of the coil and the current in the coil at different positions) have obvious effects on the thrust, specific impulse, working efficiency, service life and other indicators of the Hall thruster.

[0027] In order to design a Hall thruster with better indicators, it is necessary to perform working simulations on Hall thrusters with different parameters, and then screen out the parameters of the Hall thruster that optimize the above - mentioned indicators.

[0028] In this embodiment, a three - dimensional model of the Hall thruster is first constructed. The three - dimensional model includes an annular discharge chamber, an acceleration electrode, a coil, etc. The coil is used to generate a magnetic field in the discharge chamber to confine ions in the discharge chamber; the acceleration electrode includes a cathode and an anode, which are used to generate an electric field for accelerating ions in the discharge chamber. In addition, it also includes a particle emitter, which is used to generate propellant. The propellant includes neutral particles and electrons, and the neutral particles and electrons collide, ionize, and accelerate in the electromagnetic field of the discharge chamber.

[0029] The inner and outer diameters of the annular discharge chamber at different positions (i.e., different cross - sections) in the three - dimensional model, the position of the coil, the magnitude of the current in the coil at different positions (or the magnitude of the magnetic field), the voltage between the acceleration electrodes (or the magnitude of the electric field), the propellant flow density of the particle emitter, etc. are used as a set of parameters of the Hall thruster.

[0030] As an alternative example, multiple sets of parameters are randomly generated, and the corresponding three-dimensional model of the Hall thruster under each set of parameters can simulate the corresponding Hall thruster indicators. The indicators corresponding to different sets of parameters may be different. In this embodiment, a set of parameters when the Hall thruster indicators are optimal (for example, when the thrust and life are maximized, or when the working efficiency is maximized) is used as the optimal design parameters of the Hall thruster.

[0031] As a preferred example, a number of sets of parameters are artificially set (for example, experts set a number of sets of parameters based on experience).

[0032] Obtaining an optimal set of parameters using the indicators simulated under different sets of parameters is a conventional technique and is not the focus of this embodiment either, so this embodiment will not elaborate specifically.

[0033] In this embodiment, the three-dimensional model of the Hall thruster is divided into a number of grids (specifically, the discharge chamber of the Hall thruster is divided into a number of grids). Each grid is a cube, and the volumes of different grids are the same. In this embodiment, the volume of each grid is 0.08 cubic millimeters.

[0034] For the three-dimensional model of the Hall thruster under a given set of parameters, in this embodiment, the particle-in-cell method (PIC) is used to simulate the three-dimensional model of the Hall thruster. This method tracks the movement trajectories of a large number of particles (neutral atoms, ions, electrons), counts the collision events in each grid at each moment, and finally obtains the number of positive ions, the number of negative ions, the number of neutral particles, as well as the particle density, the average velocity of all particles, etc. in each grid. In this embodiment, the above data in each grid is spliced into a vector, and a simulation value is generated for each grid at each moment.

[0035] When the simulation process continues for a period of time (for example, after one hour), indicators such as the specific impulse, thrust, working efficiency, and life of the Hall thruster are simulated.

[0036] The particle-in-cell method (PIC) is a well-known technique, and this embodiment will not elaborate specifically.

[0037] In other embodiments, only the particle density or the average velocity of the particles can be used as the simulation value. In some other embodiments, the ratio of the sum of the number of positive ions and negative ions to the number of all particles can be recorded as the ionization rate, and the ionization rate can be used as the simulation value.

[0038] As an alternative example, each moment is one second apart. In other embodiments, each moment can be five seconds apart. It should be noted that each time a simulation is performed, the moment after all grids obtain the simulation value for the first time is recorded as the 1st moment (i.e., the starting moment).

[0039] As a preferred example, considering that during each simulation, due to the large amount of calculation (especially when the number of grids is large), it is impossible to obtain the simulation values of each grid in real time. For example, after simulating the simulation values of all grids at the current time point, the corresponding simulation values of all grids for the next time may take some time (such as 2 to 4 seconds) to be simulated. Therefore, in this example, when the simulation values are simulated for all grids, it is recorded as a moment. For example, the simulation value of each grid at the i-th moment in this example actually represents the simulation value obtained after the i-th simulation of each grid in the same simulation process.

[0040] So far, after a set of parameters of the Hall thruster is given, the simulation values of each grid at different moments can be obtained.

[0041] Step S102: Continuously change the parameters of the three-dimensional model of the Hall thruster. During the current simulation process, obtain the target area, where the simulation value differences of all grids in the target area are the smallest, and the average simulation value of all grids in the target area has the same change trend as the average simulation value of all grids in the same target area during the previous simulation process.

[0042] In order to obtain the simulation results under different parameters, this implementation continuously changes the parameters of the three-dimensional model of the Hall thruster multiple times. After each change of parameters, a full-process numerical simulation is performed. During each simulation, a simulation value is generated for each grid in the three-dimensional model of the Hall thruster at each moment.

[0043] This step takes into account that when there are many parameter sets or the number of grids is too large, a large amount of computational effort and computing time are required. Or rather, it is impossible to accurately obtain the simulation results under multiple parameter sets within a limited time (the more grids, the more accurate), which in turn leads to the inability to ensure that the simulated indicators are globally optimal (the more grids and the more parameter sets can ensure global optimality), and thus the indicators such as the thrust, specific impulse, and lifespan of the designed Hall thruster cannot meet the expectations.

[0044] Therefore, this step improves the computational speed of the simulation process by combining the adjacent two simulation processes, so that the optimal parameters of the indicators can be obtained using as many parameters and grids as possible within a limited time.

[0045] In this embodiment, the obtained target area has the same or similar change trend as the previous simulation process. Therefore, the simulation results of the previous simulation process (that is, the simulation values of each grid) can be used to assist in obtaining the simulation results during the current simulation process.

[0046] Specifically, when there is no previous simulation process, the particle-in-cell (PIC) method is used to simulate the current process.

[0047] As an example, during the current simulation process, a target area is obtained. The simulation values of all grids within the target area have the smallest difference, and the average simulation value of all grids within the target area has the same change trend as the average simulation value of all grids within the same target area during the previous simulation process. The methods included are as follows: At the current moment during the current simulation process, the simulation values of each grid are obtained. The coordinates of each grid (the coordinates of the center point of the grid) and the simulation values of the grid at each moment and the previous moment are concatenated into a vector, which is denoted as the feature point of each grid. K-Means clustering is performed on the feature points of all grids at the current moment, and the number of resulting categories is set to K1. Each resulting category is denoted as the first category. The grids within the same category are connected in terms of spatial position, and at the same time, the simulation values have the smallest difference.

[0048] To eliminate the differences in the dimensions of different quantities in the vectors corresponding to the feature points, in this embodiment, the ZCA algorithm is used to whiten the feature points of all grids.

[0049] In this embodiment, K1 is equal to the ceiling of the total number of all grids divided by 20. In other embodiments, K1 can be set to other values, and this embodiment does not make specific limitations.

[0050] The area composed of all grids within each first category is denoted as the first area. During the current simulation process, the first average value of the simulation values of all grids within the first area at each moment is obtained. During the current simulation process, the first average values at all moments before the current moment (including the current moment) form the first sequence of each first area.

[0051] Specifically, if there are less than 5 moments before the current moment, all the methods of this embodiment are not executed, but instead, the particle-in-cell (PIC) method is directly used to simulate the three-dimensional model of the Hall thruster.

[0052] During the previous simulation process, the second average value of the simulation values of all grids within each first area is obtained. During the previous simulation process, the second average values at all moments before the current moment form the second sequence of each first area.

[0053] The trend similarity between the first sequence and the second sequence of each first area is obtained. When the trend similarity is greater than or equal to the first preset threshold th1, it is determined that: the average simulation value of all grids within the first area has the same change trend as the average simulation value of all grids within the same area during the previous simulation process. At this time, the first area is denoted as the target area.

[0054] In this embodiment, th1 = 0.6 is used as an example for description. In other embodiments, th1 can be set to other values, and this embodiment does not make specific limitations.

[0055] Specifically, if the number of grids in the first region is less than 10, then the first region is deleted.

[0056] Thus, a number of target regions are obtained.

[0057] As another example, during the current simulation process, a target region is obtained, where the difference in simulation values of all grids in the target region is minimized, and the average simulation value of all grids in the target region has the same change trend as the average simulation value of all grids in the same target region during the previous simulation process. The method includes: For all first regions where the trend similarity is less than the first preset threshold th1, they are denoted as the first regions to be merged.

[0058] For any two adjacent first regions to be merged (i.e., the first regions to be merged with a common boundary), these two first regions to be merged are merged into one region, which is denoted as the first region, and then the trend similarity of this first region is calculated again using the method in the above example, which is denoted as the merged trend similarity of any two adjacent first regions to be merged.

[0059] The first regions to be merged with a merged trend similarity greater than or equal to th1 are merged, and the merged region is also denoted as the target region.

[0060] Specifically, when the merged trend similarity between any one first region to be merged and multiple other first regions to be merged is greater than or equal to th1, at this time, it is preferably to merge the two first regions to be merged with the largest merged trend similarity, and the target region is obtained.

[0061] As an example, the method for obtaining the trend similarity between the first sequence and the second sequence includes: If the simulation values in the first sequence and the second sequence are vectors, then the simulation values in the first sequence and the second sequence are PCA - dimensionally reduced to one - dimension, so that the values in the first sequence and the second sequence are one - dimensional.

[0062] The first sequence and the second sequence are respectively subjected to linear normalization processing. The DTW distance between the linearly normalized first sequence and the second sequence is denoted as x, and exp(−x) is denoted as the trend similarity, where exp() represents the exponential function with the natural constant as the base. The DTW distance is obtained by the well - known DTW algorithm.

[0063] In other embodiments, the cosine similarity between the linearly normalized first sequence and the second sequence can be denoted as the trend similarity.

[0064] Step S103: During the current simulation process, predict the first predicted value of each grid at the next moment using the time series of the simulation values of each grid within the target area. Denote the simulation values of each grid at the next moment in the same target area during the previous simulation process as the second predicted value of each grid during the current simulation process.

[0065] Since the differences between the simulation values of all grids within the target area are relatively small, the processes of particle collision, ionization, acceleration, etc. of the grids within the target area may not be interfered by the processes of particle collision, ionization, acceleration, etc. of the particles in other grids. Therefore, during the current simulation process, predict the first predicted value of each grid at the next moment using the time series of the simulation values of each grid within the target area, and this first predicted value can be used as a reference for the simulation value of each grid at the next moment.

[0066] Specifically, the time series of the simulation values of each grid refers to the time series composed of the simulation values at all moments within a preset time period (such as within 5 moments) before the current moment during the current simulation process.

[0067] As an example, the method for predicting the first predicted value of each grid at the next moment includes: Use the Kalman filter algorithm for prediction to obtain the first predicted value at the next moment. In other embodiments, the Kalman filter algorithm can be replaced by the extended Kalman filter algorithm.

[0068] As another example, the method for predicting the first predicted value of each grid at the next moment includes: For any simulation process among multiple historical simulation processes, after obtaining each grid within the first area during this simulation process, use the time series of the simulation values of each grid within the first area as a sample. When the number of obtained samples is sufficient (such as more than 300), form a data set with these samples, and use this data set to train an LSTM neural network. The input of this LSTM neural network is the time series of the simulation values, and the output is the simulation value at the next moment.

[0069] For the time series of the simulation values of each grid obtained during the current simulation process, use the trained LSTM neural network for prediction to obtain the first predicted value at the next moment.

[0070] The structure, training, and usage methods of the LSTM neural network are well-known and will not be elaborated in this embodiment.

[0071] Furthermore, since the average simulation value of all grids within the target area has the same trend of change as the average simulation value of all grids within the same target area during the previous simulation process, it indicates that the variation pattern of the grid simulation values in the current simulation process is the same as that in the previous simulation process. Therefore, the simulation values of the grids at the next moment during the previous simulation process can be used to calculate the simulation values of the grids at the next moment during the current simulation process, that is: Denote the simulation values of each grid at the next moment within the same target area during the previous simulation process as the second prediction values of each grid during the current simulation process, and the second prediction values can also be used as a reference for the simulation values of each grid at the next moment.

[0072] Step S104: During the current simulation process, obtain the evolution area, where the evolution area is adjacent to the target area and has the largest average simulation value among all areas adjacent to the target area.

[0073] Since the evolution area is adjacent to the target area and has the largest average simulation value among all areas adjacent to the target area, there is an obvious correlation between the target area and the evolution area. For example, there are particle exchanges, charge offsets, and evolutions in the particle velocity distribution due to particle collisions, ionizations, and accelerations between the target area and the evolution area.

[0074] As an example, the method for obtaining the evolution area includes: During the current simulation process, among all the first areas, obtain all the first areas adjacent to each target area, denoted as the target adjacent areas; obtain the average value of all the simulation values within each target adjacent area, that is, the average simulation value.

[0075] For the target adjacent areas of each target area, obtain the difference between the average simulation value of the target area and the average simulation value of the target adjacent area, and denote the N1 target adjacent areas with the largest differences as the evolution areas of each target area.

[0076] In this embodiment, N1 is rounded up to half of the number of target adjacent areas. In other embodiments, N1 can be set to other values, such as N1 = 1.

[0077] The difference between the average simulation values refers to the absolute value of the difference. If the simulation values are vectors, the difference between the average simulation values refers to the Euclidean distance.

[0078] Step S105: Perform weighted fusion on the first prediction values and the second prediction values of all grids within the target area to obtain the simulation values of all grids within the target area at the next moment, and the direction of the maximum distribution difference of the simulation values of all grids within the target area at the next moment is consistent with the direction of the maximum distribution difference of the simulation values of all grids within the evolution area.

[0079] The first predicted values and the second predicted values of all grids in the target area are weighted and fused to obtain the simulated values of all grids in the target area at the next moment. The process fuses the simulation results of the previous simulation process and the simulation results of the current simulation process to directly obtain the simulated values of the current simulation process at the next moment, avoiding the problems of large computational amount and long computational time when simulating a large number of grids by the particle-in-cell (PIC) method. On the other hand, it enables the accurate simulation results to be obtained in a shorter time (the more grids, the more accurate).

[0080] Furthermore, the maximum distribution difference direction of the simulated values at the next moment obtained by fusion in this embodiment, and the maximum distribution difference direction of the simulated values of all grids in the evolution area, represent the change propagation directions of the velocity, charge, ionization situation, etc. of particles in the local area.

[0081] In this embodiment, the maximum distribution difference direction of the simulated values of all grids in the target area at the next moment is consistent with the maximum distribution difference direction of the simulated values of all grids in the evolution area. For this process, from a larger range (i.e., the larger local area composed of the target area and the evolution area), the change propagation directions of the velocity, charge, ionization situation, etc. of particles will not mutate in a short time. Therefore, in this embodiment, by making the predicted change propagation direction consistent with the change propagation direction of the evolution area, the consistency of the change propagation direction in a larger range (i.e., no mutation) is ensured, and thus the accuracy of the predicted simulated values is ensured.

[0082] As an example, the specific method for obtaining the maximum distribution difference direction of the simulated values of all grids in the evolution area is as follows: For the feature points of all grids in the evolution area, the principal component directions of all feature points are obtained by using the PCA algorithm, and these principal component directions are used as the maximum distribution difference directions of the simulated values.

[0083] As another example, the specific method for obtaining the maximum distribution difference direction of the simulated values of all grids in the evolution area is as follows: For the feature points of all grids within the evolution region, the K-Means algorithm is used again to cluster all the feature points. The obtained categories are denoted as the second categories, and the number of the obtained second categories is set to N2. In this embodiment, N2 = 10 is taken as an example for description. The mean value of all feature points within each second category is denoted as the category center point of the second category. Obtain the average simulation value within each second category, obtain the difference between the average simulation values of any two second categories, obtain several pairs (such as 5 pairs) of second categories with the largest difference. The 5 pairs of vectors formed between the category center points of these 5 pairs of second categories, that is, the displacement vectors between 5 pairs of category center points. In this embodiment, it is stipulated that the direction of the displacement vector is from the feature point with a smaller modulus to the feature point with a larger modulus; the directions where these 5 pairs of vectors are located are used as the directions of the maximum distribution difference of the simulation values.

[0084] Compared with the above example, the direction of the maximum distribution difference obtained in this example is not limited to the orthogonal direction, and the number of the directions of the maximum distribution difference can be set artificially, and the calculation amount is relatively small.

[0085] The direction of the maximum distribution difference of the simulation values obtained in all examples is represented as the unit vector in this direction (denoted as the second direction).

[0086] As a preferred example, the first prediction values and the second prediction values of all grids within each target region are weighted and fused to obtain the simulation values of all grids within the target region at the next moment. The methods included are: Given several groups of fusion weights, each group of fusion weights is represented as w1 and w2, and w1 + w2 = 1. For example, 10 groups of fusion weights are given, and (w1, w2) in each group of fusion weights are respectively (0.0, 1.0), (0.1, 0.9), (0.2, 0.8), ……, (1.0, 0).

[0087] For each group of fusion weights (w1, w2), use w1 and w2 to perform weighted summation on the first prediction values and the second prediction values of all grids respectively. w1 is the weighted weight of the first prediction values of all grids, and w2 is the weighted weight of the second prediction values of all grids, and then obtain the comprehensive prediction simulation values of all grids.

[0088] Obtain the direction of the maximum distribution difference of the comprehensive prediction simulation values of all grids within the target region (the specific method is the same as the above example in this step), which is represented as the unit vector in this direction, simply denoted as the first direction. The unit vector corresponding to the direction of the maximum distribution difference of the simulation values of all grids within the evolution region is simply denoted as the second direction.

[0089] Perform KM matching between all the first directions and all the second directions to obtain a number of matching pairs. Each matching pair contains a first direction and a second direction, and a cosine similarity is calculated for the first direction and the second method of each matching pair. The sum of the absolute values of the cosine similarities of all the matching pairs obtained by the KM algorithm is the largest.

[0090] In this embodiment, the sum of the absolute values of the cosine similarities of all the matching pairs is denoted as the selection index for each group of fusion weights. Obtain a group of fusion weights with the largest selection index, and use this group of fusion weights to perform weighted summation on the first predicted values and the second predicted values of all the grids. The comprehensive predicted simulation values of all the grids obtained are used as the simulation values of all the grids in the target area at the next moment during the current simulation process.

[0091] At this time, the direction of the maximum distribution difference of the simulation values of all the grids in the target area at the next moment is the same as the direction of the maximum distribution difference of the simulation values of all the grids in the evolution area.

[0092] The computational complexity of this example is relatively small.

[0093] As another alternative example, perform weighted fusion on the first predicted values and the second predicted values of all the grids in each target area to obtain the simulation values of all the grids in the target area at the next moment. The methods included are as follows: Randomly initialize a group of fusion weights. Use this group of fusion weights to perform weighted summation on the first predicted values and the second predicted values of all the grids in each target area respectively to obtain the comprehensive predicted simulation values of all the grids in each target area. According to the direction of the maximum distribution difference of the comprehensive predicted simulation values of all the grids in the target area and the direction of the maximum distribution difference of the simulation values of all the grids in the evolution area, obtain the selection index of this group of fusion weights. Use the simulated annealing algorithm to obtain the fusion weights when the selection index is the largest. Use this group of fusion weights to perform weighted summation on the first predicted values and the second predicted values of all the grids. The comprehensive predicted simulation values of all the grids obtained are used as the simulation values of all the grids in the target area at the next moment during the current simulation process.

[0094] At this time, the direction of the maximum distribution difference of the simulation values of all the grids in the target area at the next moment is the same as the direction of the maximum distribution difference of the simulation values of all the grids in the evolution area.

[0095] Although the computational complexity of this example is large, the simulation values obtained at the next moment are more accurate.

[0096] So far, during the current simulation process, the simulation values of each grid in the target area at the next moment have been obtained, reducing the computational complexity of the simulation process. For each grid outside the target area, the particle-in-cell (PIC) method is still used to simulate the simulation values at the next moment.

[0097] Repeat the steps S102 to S105 of this embodiment so that the simulation values of all grids in the target area at the next moment are obtained by fusing the simulation values of the previous simulation process and the simulation values that have been simulated in the current simulation process. This process does not require using the simulation values of each grid by the Particle-in-Cell (PIC) method at each moment, reducing the computational amount, enabling the use of more grids and more groups of Hall thruster parameters within a limited time to accurately obtain the simulation results (i.e., the simulation values of each grid at each moment), thereby ensuring that the indicators simulated by the Hall thruster (such as specific impulse, thrust, lifetime, etc.) are more accurate, and further ensuring that the Hall thruster parameters under the optimal indicators can be obtained, so as to design a Hall thruster that meets the expectations. At the same time, in the specific process of this embodiment, the first predicted values and the second predicted values of all grids in the target area are weighted and fused, and the maximum distribution difference direction of the simulation values of all grids in the target area at the next moment is the same as the maximum distribution difference direction of the simulation values of all grids in the evolution area; the maximum distribution difference direction represents the change spread direction of the velocity, charge, ionization situation, etc. of particles in the local area; this process ensures the consistency of the change spread direction in a larger range (i.e., no mutation) by making the predicted change spread direction consistent with the change spread direction of the evolution area, and further ensures the accuracy of the predicted simulation values.

[0098] This is the end of this embodiment.

[0099] Embodiment Two: In order to further ensure the accuracy of the simulation process in this embodiment, after obtaining the simulation values of all grids in the target area at the next moment using the method in Embodiment One, then, use the Particle-in-Cell (PIC) method to simulate the simulation values of all grids at subsequent moments; after using the Particle-in-Cell (PIC) method to simulate the simulation values of all grids, use the method in Embodiment One again to obtain the simulation values of all grids in the target grid at subsequent moments, and so on until the simulation processes corresponding to all parameters are completed. This process obtains the simulation values of all grids by alternately using the Particle-in-Cell (PIC) method and the method in Embodiment One, taking into account both the computational speed and the accuracy of the simulation process.

[0100] Embodiment Three: The Hall thruster developed in this embodiment is required to have a higher working efficiency than the existing EHT-400 thruster.

[0101] The greater the propellant flow density (propellant flow rate / cross-sectional area of the discharge chamber) of the Hall thruster, the closer the plasma discharge region (i.e., the region where the particles as a whole exhibit an electrical center) is to the vicinity of the discharge chamber nozzle. At this time, the interaction between the plasma (i.e., the particle group containing charged ions) and the discharge chamber wall is smaller, the blocking effect on the ion ejection is lower, and the working efficiency of the Hall thruster is higher. As the power of the Hall thruster decreases, the propellant flow density decreases synchronously. To maintain plasma discharge, the plasma discharge region will gradually move toward the inside of the discharge chamber. At this time, the interaction between the plasma and the discharge chamber wall gradually increases, the blocking effect of the wall on the ion ejection begins to increase, and the working efficiency of the Hall thruster gradually decreases.

[0102] To suppress the phenomenon that the plasma discharge region shrinks toward the inside of the discharge chamber when the power of the Hall thruster decreases and the propellant flow rate decreases, three sets of parameters are included in the multiple sets of parameters of the Hall thruster provided in this embodiment. These three sets of parameters specifically refer to three annular discharge chambers with different structures. These three annular discharge chambers with different structures have different inner diameters and outer diameters at different positions (i.e., different cross-sections), specifically as Figure 2 、 Figure 3 、 Figure 4 shown. The difference between these three sets of parameters lies only in Figure 2 、 Figure 3 、 Figure 4 the structures shown.

[0103] Figure 3 、 Figure 4 The designed discharge chamber has the following differences compared to Figure 2 : Near the nozzle of the Hall thruster discharge chamber, a variable cross-section region is constructed so that the cross-sectional area of the discharge channel gradually decreases from the outside to the inside.

[0104] After respectively simulating all groups of parameters in Example 1, one group of parameters including the structure shown in Figure 4 has a higher working efficiency compared to the parameters including the structures shown in Figure 2 、 Figure 3 (the working efficiency is increased by 11.8 times compared to Figure 2 ).

[0105] Example 4: This embodiment provides a full-process numerical simulation system for a Hall thruster, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the full-process numerical simulation method for a Hall thruster as described in Example 1.

[0106] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A full-process numerical simulation method for Hall thrusters, characterized in that, The method includes the following steps: Changing the parameters of the Hall thruster multiple times, and performing a simulation each time after changing the parameters; during each simulation, a simulation value is generated for each grid in the three-dimensional model of the Hall thruster at each moment; During the current simulation process, obtain a target area where the simulation value differences of all grids in the target area are the smallest, and the average simulation value of all grids in the target area has the same change trend as the average simulation value of all grids in the same target area during the previous simulation process; use the time series of the simulation values of each grid in the target area to predict the first predicted value of each grid at the next moment, and record the simulation value of each grid at the next moment in the same target area during the previous simulation process as the second predicted value of each grid during the current simulation process; During the current simulation process, obtain an evolution area that is adjacent to the target area and has the largest difference from the average simulation value of the target area; Perform weighted fusion on the first predicted values and the second predicted values of all grids in the target area to obtain the simulation values of all grids in the target area at the next moment, and the direction of the maximum distribution difference of the simulation values of all grids in the target area at the next moment is consistent with the direction of the maximum distribution difference of the simulation values of all grids in the evolution area; 2. The full-process numerical simulation method for a Hall thruster according to claim 1, wherein The obtaining of the target area, where the simulation value differences of all grids in the target area are the smallest, and the average simulation value of all grids in the target area has the same change trend as the average simulation value of all grids in the same target area during the previous simulation process, includes the following specific steps: At the current moment in the current simulation process, perform K-Means clustering on the feature points of all grids at the current moment to obtain several first categories; the area composed of all grids in each first category is recorded as the first area; during the current simulation process, obtain the first average value of the simulation values of all grids in the first area at each moment; during the current simulation process, the first average values at all moments before the current moment form the first sequence of each first area; During the previous simulation process, obtain the second average value of the simulation values of all grids in each first area. During the previous simulation process, the second average values at all moments before the current moment form the second sequence of each first area; Obtain the trend similarity between the first sequence and the second sequence of each first area. When the trend similarity is greater than or equal to the first preset threshold, it is determined that the average simulation value of all grids in the first area has the same change trend as the average simulation value of all grids in the same area during the previous simulation process, and the first area is recorded as the target area; 3. The full-process numerical simulation method for Hall thrusters according to claim 2, wherein The obtaining of the evolution area, where the evolution area is adjacent to the target area and has the largest difference from the average simulation value of the target area, includes the following specific steps: During the current simulation process, among all the first areas, obtain all the first areas adjacent to each target area, which are recorded as the target adjacent areas; obtain the average simulation value of all grids in each target adjacent area; Obtain the difference between the average simulation value of the target area and the average simulation value of the target adjacent areas, and record several target adjacent areas with the largest difference as the evolution areas of each target area.

4. The full-process numerical simulation method for Hall thrusters according to claim 1, characterized in that The specific steps for weighted fusion of the first prediction values and the second prediction values of all grids in the target area to obtain the simulation values of all grids in the target area at the next moment are as follows: Preset several groups of fusion weights. For each group of fusion weights (w1, w2), use w1 and w2 to perform weighted summation on the first prediction values and the second prediction values of all grids in the target area respectively. w1 is the weighted weight of the first prediction values of all grids, and w2 is the weighted weight of the second prediction values of all grids, to obtain the comprehensive prediction simulation values of all grids. Obtain the maximum distribution difference direction of the comprehensive prediction simulation values of all grids in the target area, briefly denoted as the first direction; the maximum distribution difference direction of the simulation values of all grids in the evolution area is briefly denoted as the second direction; obtain the selection index of each group of fusion weights using the similarity between the first direction and the second direction; obtain the group of fusion weights with the largest selection index, and use this group of fusion weights to perform weighted summation on the first prediction values and the second prediction values of all grids. The obtained comprehensive prediction simulation values of all grids are used as the simulation values of all grids in the target area at the next moment during the current simulation process.

5. The full-process numerical simulation method for Hall thrusters according to claim 1, characterized in that The specific steps for obtaining the maximum distribution difference direction of the simulation values of all grids in the evolution area are as follows: Use the K-Means algorithm to cluster the feature points of all grids in the evolution area to obtain several second categories, obtain the average simulation value within each second category, obtain the difference between the average simulation values of any two second categories, obtain several pairs of second categories with the largest difference, and the direction where the displacement vector is located between the category center points of several pairs of second categories is used as the maximum distribution difference direction of the simulation values of all grids in the evolution area.

6. The full-process numerical simulation method for Hall thrusters according to claim 2, characterized in that The specific steps for obtaining the trend similarity between the first sequence and the second sequence of each first area are as follows: Perform linear normalization processing on the first sequence and the second sequence respectively. Denote the DTW distance between the linearly normalized first sequence and the second sequence as x, and denote exp(-x) as the trend similarity, where exp() represents the exponential function with the natural constant as the base.

7. The full-process numerical simulation method for Hall thrusters according to claim 2 or 5, characterized in that The feature points of the grid are: the vector formed by splicing the center point coordinates of the grid and the simulation values of the grid at adjacent moments.

8. The full-process numerical simulation method for Hall thrusters according to claim 4, characterized in that The specific steps for obtaining the selection index of each group of fusion weights using the similarity between the first direction and the second direction are as follows: Perform KM matching between all the first directions and all the second directions to obtain several matching pairs. Each matching pair contains a first direction and a second direction, and calculate the cosine similarity between the first direction and the second direction of each matching pair; the sum of the cosine similarities of all matching pairs is denoted as the selection index of each group of fusion weights.

9. The full-process numerical simulation method for Hall thrusters according to claim 8, wherein The parameters include the inner diameter and outer diameter of the annular discharge chamber of the Hall thruster at different positions, the current magnitude of the coil, the voltage between the acceleration electrodes, and the propellant flow density of the particle emitter.

10. A full-process numerical simulation system for a Hall thruster, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the steps of the full-process numerical simulation method for a Hall thruster according to any one of claims 1 to 9 are implemented.

Images