Real-time prediction method for ion electric propulsion beam flicker based on Bayesian network

Abstract

The invention relates to the technical field of space electric propulsion, in particular to an ion electric propulsion beam flicker real-time prediction method based on a Bayesian network, and the method comprises the steps: constructing a Bayesian network model of an ion electric thruster, updating the node probability in combination with monitoring data, and predicting the occurrence probability of beam flicker in real time. Therefore, beam flicker is early warned in advance, and a pre-suppression measure is given. The method adapts to the characteristic of random generation of ion electric propulsion beam flicker, can give an early warning for the generation of the beam flicker, reduces the failure risk of an ion electric propulsion product, and improves the robustness and reliability of the ion electric propulsion product.

Description

Technical Field

[0001] The present application relates to the field of space electric propulsion technology, and in particular to a real-time prediction method for ion electric propulsion beam scintillation based on a Bayesian network. Background Art

[0002] Ion electric propulsion is an important power technology in the space field. Its principle is to ionize the gas propellant in a discharge chamber to generate plasma, and then apply a voltage difference to the two grids downstream of the discharge chamber to draw out the plasma through electrostatic acceleration to generate thrust.

[0003] Because the gate voltage difference is typically around 2000V, and the gate spacing is only on the millimeter scale, the coupling of multiple factors, including strong electric fields, the inter-gate gas environment, the gate surface conditions, and unwanted particles, can cause gate breakdown. This can manifest as a macroscopic beam instability, known as beam flicker. Frequent beam flicker can adversely affect the lifespan and reliability of ion electric propulsion products, and in severe cases, can damage the product or even lead to the failure of the entire satellite mission.

[0004] Therefore, it is necessary to take measures to suppress beam flicker. However, due to the strong coupling and randomness of beam flicker, the current "post-processing" method is used, that is, turning off the power when beam flicker occurs, or evaporating, ablating, or baking the excess bridging materials after the beam flicker occurs. It is not possible to pre-process the beam flicker. These processing methods have a significant impact on the normal operation of the thruster, causing thrust interruption and thrust loss. Summary of the Invention

[0005] The present application provides a real-time prediction method for ion electric propulsion beam flicker based on a Bayesian network, which can update node probabilities in combination with monitoring data and predict the probability of beam flicker in real time, thereby achieving early warning of beam flicker and providing pre-emptive suppression measures.

[0006] In order to achieve the above-mentioned purpose, the present application provides a real-time prediction method for ion electric propulsion beam flicker based on a Bayesian network, comprising the following steps: Step 1: constructing a Bayesian network model of an ion electric thruster and giving the causal relationship between node variables; Step 2: starting the ion electric thruster and continuously collecting monitorable parameter data; selecting a certain moment, recorded as moment k, and calculating the root node probability at moment k in combination with the monitoring data; Step 3: taking the root node as a priori probability, calculating the conditional probability of the leaf node, and forming a conditional probability table of the entire Bayesian network at moment k; Step 4: taking the leaf node as the priori probability, calculating the conditional probability of the leaf node, and forming a conditional probability table of the entire Bayesian network at moment k; Step 5: point as the prior probability, calculate the conditional probability of the top node, that is, the probability of beam flickering; Step 5: Set the acceptable probability threshold of beam flickering. If it is lower than the threshold, it is considered to have no impact, and if it is higher than the threshold, it is unacceptable, and feedback is given to the ion electric thruster power control and processing unit to guide the thruster state optimization; Step 6: Let k = k + 1, that is, change the time to k + 1, repeat steps 3-5, update the probabilities of the root node, leaf node and top node, recalculate the probability of beam flickering, and judge the possibility of occurrence; Step 7: Shut down the ion electric thruster and the mission ends.

[0007] Furthermore, the Bayesian network model of the ion electric thruster is constructed according to the idea of ​​dividing the beam scintillation mechanism into stages of "influencing factors-induced breakdown-beam scintillation occurrence".

[0008] Furthermore, the constructed Bayesian network model of the ion electric thruster includes a bottom layer, a middle layer and a top layer, wherein: the bottom layer is a factor layer, with sputtering deposition conditions, surface roughness, excess matter conditions, inter-gate gas pressure conditions, inter-gate plasma density and insulating material state as root nodes, that is, the influencing factors inducing beam flicker are the root nodes; the middle layer is a breakdown layer, with gate surface field-induced emission breakthrough, excess matter breakdown, inter-gate gas breakdown, inter-gate plasma conduction and insulating material surface arcing as leaf nodes, that is, the breakdown phenomenon is the leaf node; the top layer is a result layer, with beam flicker occurrence as the only node, that is, the beam flicker occurrence is the top node.

[0009] Furthermore, in step 2, when calculating the root node probability, the Bayesian formula is used to calculate the root node state transition probability, and the current root node probability is obtained by combining the observation parameter correction; among them, the prior probability in the calculation formula is given according to experimental statistical data, and the posterior probability is given by expert experience combined with mechanism analysis.

[0010] Furthermore, in step 3, when calculating the conditional probability of the leaf node, it is calculated according to the Bayesian formula; wherein, one of the prior probabilities in the calculation formula is the root node probability calculated in step 2, the other prior probability is given according to experimental statistical data, and the posterior probability is obtained by using the simulation result statistics method.

[0011] Furthermore, in step 4, when calculating the conditional probability of the top node, it is calculated according to the Bayesian formula; wherein, one of the prior probabilities in the calculation formula is the conditional probability of the leaf node calculated in step 3, the other prior probability is obtained through statistics of ion electric propulsion life test data, and the posterior probability is obtained by using the statistical method of simulation results.

[0012] Furthermore, in step 5, the acceptable probability threshold of beam scintillation is 0.5.

[0013] Furthermore, when the probability of beam flicker is between 0.5-0.65, it is a minor warning, and the thruster is subject to system-level inspection and optimization; when the probability of beam flicker is between 0.65-0.85, it is a moderate warning, and the thruster is subject to key component-level inspection and optimization; when the probability of beam flicker is above 0.85, it is a severe warning, and the thruster as a whole is subject to detailed inspection and optimization.

[0014] The present application provides a Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation, which has the following beneficial effects:

[0015] This application can combine monitoring data to update node probabilities and predict the probability of beam flickering in real time, thereby achieving early warning of beam flickering and providing pre-emptive suppression measures; this application adapts to the characteristics of random occurrence of ion electric propulsion beam flickering, and can provide early warning of beam flickering occurrence, reduce the risk of failure of ion electric propulsion products, and improve the robustness and reliability of ion electric propulsion products. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings that constitute part of this application are used to provide a further understanding of this application and make other features, objects and advantages of this application more apparent. The illustrative embodiment drawings of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. In the drawings:

[0017] Figure 1 1 is a flow chart of a Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to an embodiment of the present application;

[0018] Figure 2 is a schematic diagram of a Bayesian network model provided according to an embodiment of the present application;

[0019] Figure 3 This is a schematic diagram of the root node probability calculation provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings 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 in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] In this application, terms such as "upper," "lower," "left," "right," "front," "back," "top," "bottom," "inner," "outer," "center," "vertical," "horizontal," "transverse," and "longitudinal" indicate positions or locations based on the positions or locations shown in the accompanying drawings. These terms are primarily intended to better describe this application and its embodiments and are not intended to limit the devices, elements, or components indicated to having a specific orientation, or to being constructed or operated in a specific orientation.

[0023] Furthermore, some of the above terms may be used to express other meanings besides indicating a position or location. For example, the term "on" may also be used to express a dependency or connection in certain circumstances. Those skilled in the art will understand the specific meanings of these terms in this application based on the specific circumstances.

[0024] Additionally, the term "plurality" shall mean two or more.

[0025] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] like Figure 1 As shown, the present application provides a real-time prediction method for ion electric propulsion beam scintillation based on a Bayesian network, comprising the following steps:

[0027] Step 1: Construct a Bayesian network model of the ion electric thruster and give the causal relationship between node variables;

[0028] The Bayesian network model of the ion electric thruster is constructed according to the idea of ​​dividing the beam scintillation mechanism into stages of "influencing factors-induced breakdown-beam scintillation occurrence". That is, when the influencing factors are in the induced risk range, they will induce breakdown with a certain probability. After the breakdown occurs, beam scintillation will be induced with a certain probability. The causal relationship between the node variables of the Bayesian network is determined according to the beam scintillation inducing mechanism.

[0029] Specifically, such as Figure 2 As shown in the figure, the constructed Bayesian network model of ion electric thruster consists of three layers, the bottom node variables are marked as v i , i represents the root node number), the middle layer (the node variable is marked as u j , j represents the leaf node number) and the top layer (node ​​variable is marked as T), where: the bottom layer is the factor layer, with sputtering deposition conditions (v1), surface roughness (v2), large-sized excess particles (v3), excess particle overlap conditions (v4), inter-gate gas pressure conditions (v5), excess particle evaporation / ionization conditions (v6), inter-gate plasma density (v7), and insulating material state (v8) as root nodes; the middle layer is the breakdown layer, with gate surface field emission conditions (u1), excess particle breakdown (u2), inter-gate gas pressure breakdown (u3), inter-gate plasma conduction (u4), and insulating material surface arcing (u5) as leaf nodes; the top layer is the result layer, with beam flicker occurrence (T) as the only node.

[0030] Step 2: Start the ion electric thruster and continuously collect the data of the monitored parameters; select a certain moment, recorded as moment k, and calculate the root node probability at moment k based on the monitored data;

[0031] Specifically, the state of the node variable at time k (k = 0, 1, ...) is represented as Γ k According to engineering experience, the node variable state can be regarded as a Markov process; the state Γ value is:

[0032]

[0033] Then, if Figure 3 As shown, the Bayesian formula is used to calculate the state transition probability of the root node, and the observation parameter M k Compared with the state v at the previous moment i,k-1 Together they determine the state of the next moment:

[0034]

[0035] Among them, P(v i,k =G k |v i,k-1 =Gk-1 ,M k ) represents the root node v i From Γ k-1 The state transition is Γ k , when k=0, P(v i,k =G k |v i,k-1 =G k-1 ,M k )=P(v i,k =G k ); P(v i,k-1 =G k-1 ) represents the k-1 moment v i In state Γ i The prior probability (P(v i,k =G k ) Similarly), the prior probability can be given based on the statistical data of the beam scintillation test; when the states at two moments are consistent, P(v i,k-1 =G k-1 ) and P(v i,k =G k ) are the same; P(v i,k-1 =G k-1 ,M k |v i,k =G k ) means v i,k =G k Under the conditions that have already occurred i,k-1 =G k-1 And the observation parameter is M k The probability of scintillation is given by expert experience combined with analysis of the beam scintillation mechanism.

[0036] Step 3: Take the root node as the prior probability, calculate the conditional probability of the leaf nodes, and form the conditional probability table of the entire Bayesian network at time k;

[0037] Specifically, the leaf node u at time k i,k The conditional probability of is solved according to the following formula:

[0038]

[0039] Among them, P(v j,k ) is the root node probability calculated in step 2; the leaf node prior probability P(u i,k ) can be given based on the statistical data of beam scintillation special test; the posterior probability P(v r,k |u i,k ) is obtained by using the simulation result statistics method, that is, in the gate beam flicker simulation model, the statistics of various types of breakdown are based on the root node v i,k The frequency of the cause is used to calculate the probability, that is:

[0040]

[0041] in, To break down u i Frequency of occurrence, For v r The frequency of the cause.

[0042] Step 4: Take the leaf node as the prior probability and calculate the conditional probability of the top node, that is, the probability of beam scintillation;

[0043] Specifically, the conditional probability of the top node at time k (i.e., the probability of beam scintillation) is solved according to the following formula:

[0044]

[0045] Among them, P(u j,k ) is the root node probability calculated in step 3; the top node prior probability P(T k ) can be easily obtained by counting the number of beam flashes from the ion electric propulsion life test data; the posterior probability P(u i,k |T k ) is obtained by using the statistical method of simulation results, that is, in the grid beam flicker simulation model, the statistical beam flicker is based on the number of leaf nodes u i,k The frequency of the cause is used to calculate the probability, that is:

[0046]

[0047] Among them, N T is the frequency of beam scintillation, For u i,k The frequency of the cause.

[0048] Step 5: Set an acceptable probability threshold for beam scintillation. If the probability is below the threshold, it is considered to have no impact, and if it is above the threshold, it is unacceptable. Feedback is sent to the ion electric thruster power control and processing unit to guide thruster state optimization.

[0049] Specifically, based on engineering experience, the threshold of the acceptable probability of beam flicker is set at 0.5. If the probability of beam flicker is lower than 0.5, it is considered to have no impact. When the probability of beam flicker is between 0.5-0.65, it is a minor warning, and the thruster is subject to system-level inspection and optimization. When the probability of beam flicker is between 0.65-0.85, it is a moderate warning, and the thruster is subject to key component-level inspection and optimization. When the probability of beam flicker is above 0.85, it is a severe warning, and the thruster as a whole is subject to detailed inspection and optimization.

[0050] Step 6: Set k = k + 1, that is, change the time to time k + 1, repeat steps 3-5, update the probabilities of the root node, leaf nodes, and top node, recalculate the probability of beam scintillation, and determine the likelihood of occurrence;

[0051] Step 7: The ion thrusters are shut down and the mission is complete.

[0052] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A real-time prediction method for ion electric propulsion beam scintillation based on Bayesian network, characterized in that: The steps include: Step 1: Construct a Bayesian network model of the ion electric thruster and give the causal relationship between node variables; Step 2: Start the ion electric thruster and continuously collect the data of the monitored parameters; select a certain moment, recorded as moment k, and calculate the root node probability at moment k based on the monitored data; Step 3: Take the root node as the prior probability, calculate the conditional probability of the leaf nodes, and form the conditional probability table of the entire Bayesian network at time k; Step 4: Take the leaf node as the prior probability and calculate the conditional probability of the top node, that is, the probability of beam scintillation; Step 5: Set an acceptable probability threshold for beam scintillation. If the probability is below the threshold, it is considered to have no impact, and if it is above the threshold, it is unacceptable. Feedback is sent to the ion electric thruster power control and processing unit to guide thruster state optimization. Step 6: Set k = k + 1, that is, change the time to time k + 1, repeat steps 3-5, update the probabilities of the root node, leaf nodes, and top node, recalculate the probability of beam scintillation, and determine the likelihood of occurrence; Step 7: The ion thrusters are shut down and the mission is complete.

2. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 1, characterized in that: The Bayesian network model of ion electric thrusters is constructed according to the idea of ​​dividing the beam scintillation mechanism into stages: "influencing factors-induced breakdown-beam scintillation occurrence".

3. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 2, characterized in that: The constructed Bayesian network model of ion electric thruster includes the bottom layer, the middle layer and the top layer, among which: The bottom layer is a factor layer, with sputtering deposition conditions, surface roughness, excess material conditions, inter-grid gas pressure conditions, inter-grid plasma density, and insulating material conditions as root nodes, that is, the factors influencing the induction of beam flickering as root nodes; The intermediate layer is a breakdown layer, and the leaf nodes are field-induced emission breakdown on the gate surface, excess material breakdown, inter-gate gas breakdown, inter-gate plasma conduction, and arcing on the insulating material surface, that is, the breakdown phenomenon is the leaf node; The top layer is a result layer, which takes the occurrence of beam scintillation as the only node, that is, the occurrence of beam scintillation as the top node.

4. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 3, characterized in that: In step 2, when calculating the root node probability, the Bayesian formula is used to calculate the root node state transition probability, and the current root node probability is obtained by combining the observation parameter correction; among them, the prior probability in the calculation formula is given according to the experimental statistical data, and the posterior probability is given by expert experience combined with mechanism analysis.

5. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 4, characterized in that: In step 3, when calculating the conditional probability of the leaf node, it is calculated according to the Bayesian formula; among them, one of the prior probabilities in the calculation formula is the root node probability calculated in step 2, the other prior probability is given according to the experimental statistical data, and the posterior probability is obtained by the simulation result statistics method.

6. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 5, characterized in that: In step 4, when calculating the conditional probability of the top node, it is calculated according to the Bayesian formula; among them, one of the prior probabilities in the calculation formula is the conditional probability of the leaf node calculated in step 3, the other prior probability is obtained through the statistics of the ion electric propulsion life test data, and the posterior probability is obtained by the statistical method of simulation results.

7. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 6, characterized in that: In step 5, the acceptable probability threshold of beam scintillation is 0.

5.

8. The Bayesian network-based real-time prediction method for ion electric propulsion beam scintillation according to claim 7, characterized in that: When the probability of beam flicker is between 0.5-0.65, it is a minor warning, and the thruster is subject to system-level inspection and optimization; when the probability of beam flicker is between 0.65-0.85, it is a moderate warning, and the thruster is subject to key component-level inspection and optimization; when the probability of beam flicker is above 0.85, it is a severe warning, and the thruster as a whole is subject to detailed inspection and optimization.

Images