Fault detection method for test launching stage of carrier rocket

Abstract

The invention discloses a fault detection method for a carrier rocket test launch stage, which comprises the following steps of: firstly, constructing a carrier rocket test launch process framework with a four-layer tree structure, then, carrying out directed graph description on each test item, carefully distributing graphic code names for test quantities, detailedly assigning a plurality of states and orderly numbering, and finally, carrying out fault detection on the test quantities. Nodes are connected strictly according to rules such as direct causal and sequential relation, and a clear directed graph is drawn. On the basis, five basic rules are derived and ingeniously fused into a complete fault detection rule. Finally, based on a rule-based fault diagnosis expert system, the rules are accurately applied to carry out fault discrimination on the whole test launching process of the carrier rocket. According to the graphical means, fault detection rule writing is tedious, visual and easy to understand, the complexity is greatly reduced, the efficiency is improved, it is practically guaranteed that a rocket launching task is smoothly propelled, and the reliability is remarkably enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field related to the test launch of launch vehicles, and particularly relates to a fault detection method for the test launch stage of launch vehicles. Background Art

[0002] Launch vehicles are expensive and have special missions, with extremely high reliability requirements. Whether in the development stage or the operation stage, the fault detection of launch vehicles is an important means to ensure their reliability and is given special attention. In particular, the fault detection in the test launch stage of launch vehicles is of great significance for the successful launch of launch vehicles.

[0003] At present, the rule-based fault diagnosis expert system is an effective way to solve the fault detection in the test launch stage of launch vehicles. However, the existing technology has problems such as complex rule writing process, large workload, long working cycle, easy to make mistakes and difficult to check, engineering and technical personnel cannot complete the rule writing work independently, the rules are difficult to understand intuitively, and it is difficult to realize rule reuse. Summary of the Invention

[0004] The purpose of the present invention is to provide a fault detection method for the test launch stage of launch vehicles to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A fault detection method for the test launch stage of launch vehicles, characterized in that the detection method steps are as follows:

[0007] Establish a tree-shaped test launch process framework for launch vehicles. The test launch process framework for launch vehicles is a four-layer tree structure, which is successively the launch vehicle model, each test time period in the test launch stage of the launch vehicle, the subsystems involved in the test in each test time period, and each test item under the subsystem from the top layer to the bottom layer.

[0008] Describe each test item with a directed graph to obtain the directed graph corresponding to each test item, specifically including:

[0009] For each test item, specify the command quantity to start the test item. There are multiple test quantities under each test item. The test quantities are divided into three types: command quantity, switch quantity, and analog quantity. The command quantity refers to the operation instruction for the test item input externally, including two states: "start" and "good";

[0010] For each test quantity under each test item, assign a graphical code name, with the initial value of the graphical code name being 0; specify various states of each test quantity in the test item theoretically, and assign natural number numbers in the order of appearance of the states; for the states of analog quantities after numbering, respectively give the basis for judging whether it is in this state according to the test value.

[0011] Take the states of each test quantity after numbering as the nodes of a directed graph. Between the nodes, connect them with directed arcs that can carry delay time to form the directed graph of the test item. When using directed arcs to connect nodes, it is required that there is a direct causal relationship or a direct sequential relationship between the start point and the end point of the directed arc, and the node corresponding to the analog quantity state cannot be used as the start point of the directed arc. For the direct causal relationship, the start point is the cause and the end point is the result; for the direct sequential relationship, the start point is earlier and the end point is later.

[0012] Preferably, according to the directed graph, generate five basic rules:

[0013] For each directed arc without delay time in the directed graph, the generated basic rule is that if the value of the graphical code name of the test quantity corresponding to the start point of the directed arc is equal to the number corresponding to the state of the start point of the directed arc, then the value of the graphical code name of the test quantity corresponding to the end point of the directed arc is assigned the number corresponding to the state of the end point of the directed arc;

[0014] For each directed arc with delay time in the directed graph, generate two basic rules, which are respectively: if the value of the graphical code name of the test quantity corresponding to the start point of the directed arc is equal to the number corresponding to the state of the start point of the directed arc, then set a timer and start timing; if the timing result of this timer reaches the delay time carried by the directed arc, then the value of the graphical code name of the test quantity corresponding to the end point of the directed arc is assigned the number corresponding to the state of the end point of the directed arc;

[0015] For the "start" state node of the start instruction quantity in the directed graph, the generated basic rule is that if the test value of the start instruction quantity is equal to the theoretical value of the start instruction quantity at this node, then the value of the graphical code name of the start instruction quantity is assigned the number of this node;

[0016] For each node corresponding to a switch quantity in the directed graph, generate two basic rules in sequence: if the test value of the switch quantity is equal to the theoretical value of the switch quantity at this node, then the value of the graphical code name of the switch quantity is assigned the number of this node; if the value of the graphical code name of the switch quantity is equal to the number of this node, then the expected value of the test value of the switch quantity is assigned the theoretical value of the switch quantity at this node;

[0017] For each node corresponding to an analog quantity in the directed graph, the basic rule generated is that if the value of the graphic code name of the analog quantity is equal to the number of the node, the test value of the analog quantity should meet the basis for judging the test value of the node;

[0018] Preferably, the five basic rules generated are integrated into a fault detection rule, including: If the antecedent of a basic rule contains the "start" state of the start instruction quantity of the test item, this basic rule is called a starting point constraint basic rule, and other basic rules are called non-starting point constraint basic rules; If the antecedent of a non-starting point constraint basic rule is the same as the consequent of a starting point constraint basic rule, then replace the antecedent of the non-starting point constraint basic rule with the antecedent of the starting point constraint basic rule, and the consequent of the non-starting point constraint basic rule remains unchanged; After processing, if the consequents of two rules are the same, then these two rules are integrated into one rule: If the antecedents of both rules hold, then the common consequent of the two rules holds.

[0019] Preferably, according to the fault detection rules in the test launch phase of the launch vehicle generated in the above steps, a rule-based fault diagnosis expert system is used to detect faults in the test launch phase of the launch vehicle.

[0020] During the process of constructing the directed graph, a dynamic adaptive adjustment algorithm is also introduced. When unexpected working conditions occur during the test or the external environment changes significantly, this algorithm can automatically adjust key parameters such as the node connection relationship and delay time of the directed graph according to the preset threshold range and the real-time collected data. For example, if a rocket fuel filling test is carried out in cold weather, the low temperature may affect the flow rate of the fuel and the response characteristics of the pressure sensor. At this time, the algorithm will appropriately extend the delay time of the directed arc related to the fuel flow according to the data feedback by the temperature sensor to ensure that the fault detection rule still accurately adapts to the actual working conditions and improve the robustness of the system.

[0021] During the operation stage of the rule-based fault diagnosis expert system, a multi-source data fusion verification module is added. This module not only collects the test data of each system of the launch vehicle itself, but also integrates the data from the environmental monitoring equipment around the launch site. By cross-verifying the multi-source data, the risk of misjudgment is further reduced. For example, when it is detected that there is an abnormal voltage fluctuation in a certain electrical system of the rocket, and at the same time a geomagnetic storm occurs near the launch site, the multi-source data fusion verification module will comprehensively analyze the correlation between the two and judge whether the voltage fluctuation is due to an electrical fault in the rocket itself or interference from the geomagnetic storm, providing a more comprehensive and accurate basis for fault diagnosis.

[0022] Compared with the prior art, the present invention provides a fault detection method for the test launch phase of a launch vehicle, having the following beneficial effects:

[0023] The graphical way of writing fault detection rules is more intuitive, reducing the complexity of rule writing. It can clearly express the logical and hierarchical relationships in the testing process, facilitating rule writers to express their understanding of the testing process in a graphical description, and also facilitating rule users to understand the rules and rule checkers to check the rules.

[0024] The graphical way of writing fault detection rules is relatively simple, reducing the difficulty of rule writing. It is easier to be understood by rule writers, facilitating the development of rule writing work, improving work efficiency, and also facilitating the implementation of advanced rule management requirements such as rule reuse. Brief Description of the Drawings

[0025] Figure 1 It is a flowchart of the detection method of the present invention. Detailed Embodiments

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0027] The present invention provides a Figure 1 fault detection method for the test launch stage of a launch vehicle as shown below. The steps of the detection method are as follows:

[0028] Establish a tree-shaped framework for the test launch process of the launch vehicle. The framework for the test launch process of the launch vehicle is a four-layer tree structure, which from top to bottom are the launch vehicle model, each test time period in the test launch stage of the launch vehicle, the subsystems involved in the test in each test time period, and each test item under the subsystem.

[0029] Describe each test item with a directed graph to obtain the directed graph corresponding to each test item, which specifically includes:

[0030] For each test item, specify the command quantity to start the test item. There are multiple test quantities under each test item, and the test quantities are divided into three types: command quantity, switch quantity, and analog quantity. The command quantity refers to the operation command for the test item input externally, including two states: "start" and "good".

[0031] For each test quantity under each test item, assign a graphical code name to it, and the initial value of the graphical code name is 0; specify the various states of each test quantity in the test item theoretically, and assign natural number numbers in the order of the appearance of the states; for the states of the analog quantity after numbering, respectively give the basis for judging whether it is in this state according to the test value.

[0032] Taking the status of each numbered test quantity as the nodes of a directed graph, between the nodes, a directed graph of test items is formed by connecting with directed arcs that can carry delay time. When using directed arcs to connect nodes, it is satisfied that there is a direct causal relationship or a direct sequential relationship between the starting point and the ending point of the directed arc, and the node corresponding to the analog quantity status cannot be used as the starting point of the directed arc. For the direct causal relationship, the starting point is the cause and the ending point is the result; for the direct sequential relationship, the starting point is earlier and the ending point is later.

[0033] According to the directed graph, five basic rules are generated:

[0034] For each directed arc without delay time in the directed graph, the generated basic rule is that if the value of the graphic code name of the test quantity corresponding to the starting point of the directed arc is equal to the number corresponding to the status of the starting point of the directed arc, then the value of the graphic code name of the test quantity corresponding to the ending point of the directed arc is assigned the number corresponding to the status of the ending point of the directed arc;

[0035] For each directed arc with delay time in the directed graph, two basic rules are generated, which are respectively: if the value of the graphic code name of the test quantity corresponding to the starting point of the directed arc is equal to the number corresponding to the status of the starting point of the directed arc, then set a timer and start timing; if the timing result of this timer reaches the delay time carried by the directed arc, then the value of the graphic code name of the test quantity corresponding to the ending point of the directed arc is assigned the number corresponding to the status of the ending point of the directed arc;

[0036] For the "start" status node of the start instruction quantity in the directed graph, the generated basic rule is that if the test value of the start instruction quantity is equal to the theoretical value of the start instruction quantity at this node, then the value of the graphic code name of the start instruction quantity is assigned the number of this node;

[0037] For each node corresponding to a switch quantity in the directed graph, two basic rules are generated in sequence: if the test value of the switch quantity is equal to the theoretical value of the switch quantity at this node, then the value of the graphic code name of the switch quantity is assigned the number of this node; if the value of the graphic code name of the switch quantity is equal to the number of this node, then the expected value of the test value of the switch quantity is assigned the theoretical value of the switch quantity at this node;

[0038] For each node corresponding to an analog quantity in the directed graph, the generated basic rule is that if the value of the graphic code name of the analog quantity is equal to the number of this node, then the test value of the analog quantity should meet the test value judgment basis of this node;

[0039] The five generated basic rules are integrated into fault detection rules, including: If the antecedent of a basic rule contains the "start" state of the start instruction quantity of a test item, the basic rule is called a starting point constraint basic rule, and other basic rules are called non-starting point constraint basic rules; If the antecedent of a non-starting point constraint basic rule is the same as the consequent of a starting point constraint basic rule, then replace the antecedent of the non-starting point constraint basic rule with the antecedent of the starting point constraint basic rule, and the consequent of the non-starting point constraint basic rule remains unchanged; After processing, if the consequents of two rules are the same, then these two rules are integrated into one rule: If the antecedents of both rules hold, then the common consequent of the two rules holds.

[0040] According to the fault detection rules in the test launch stage of the launch vehicle generated in the above steps, the faults in the test launch stage of the launch vehicle are detected by a rule-based fault diagnosis expert system.

[0041] During the process of constructing the directed graph, a dynamic adaptive adjustment algorithm is also introduced. When unexpected working conditions occur during the test or significant changes occur in the external environment, the algorithm can automatically adjust key parameters such as the node connection relationship and delay time of the directed graph according to the preset threshold range and the real-time collected data. For example, if a rocket fuel filling test is carried out in cold weather, the low temperature may affect the flow rate of the fuel and the response characteristics of the pressure sensor. At this time, the algorithm will appropriately extend the delay time of the directed arc related to the fuel flow according to the data fed back by the temperature sensor to ensure that the fault detection rules still accurately adapt to the actual working conditions and improve the robustness of the system.

[0042] In the operation stage of the rule-based fault diagnosis expert system, a multi-source data fusion verification module is added. This module not only collects the test data of each system of the launch vehicle itself, but also integrates the data from the environmental monitoring equipment around the launch site. By cross-verifying the multi-source data, the risk of misjudgment is further reduced. For example, when it is detected that there is an abnormal voltage fluctuation in a certain electrical system of the rocket, and at the same time a geomagnetic storm occurs near the launch site, the multi-source data fusion verification module will comprehensively analyze the correlation between the two to determine whether the voltage fluctuation is due to an electrical fault in the rocket itself or interference from the geomagnetic storm, providing a more comprehensive and accurate basis for fault diagnosis.

[0043] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A fault detection method for a launch vehicle test launch phase, characterized in that: The detection method steps are as follows: A tree-like launch vehicle test and launch process framework is established. The launch vehicle test and launch process framework has a four-layer tree structure. From the top layer to the bottom layer, they are the launch vehicle model, the test time periods of the launch vehicle test and launch phase, the subsystems involved in the test in each test time period, and the test items under the subsystem.

2. A fault detection method for a launch vehicle test launch phase according to claim 1, characterized in that: Each test item is described as a directed graph, and a directed graph corresponding to each test item is obtained, which specifically includes: For each test item, specify the instruction quantity for starting the test item. Each test item has multiple test quantities, which are divided into three types: instruction quantity, switch quantity and analog quantity. The instruction quantity refers to the operation instruction of the test item input from the outside, including two states of "start" and "good"; For each test quantity under each test item, a graphic code name is assigned to it, and the initial value of the graphic code name is 0; the various states of each test quantity in the test item are theoretically specified, and natural number numbers are assigned according to the order in which the states appear; for the states of the analog quantities after the numbers, the basis for judging whether they are in the state according to the test value is given respectively; The state of each numbered test quantity is a node of a directed graph. The nodes are connected with directed arcs with delayed time to form a directed graph of the test items. When using directed arcs to connect nodes, the starting point and the end point of the directed arc must be in a direct causal relationship or a direct sequential relationship, and the node corresponding to the analog quantity state cannot be used as the starting point of the directed arc. For a direct causal relationship, the starting point is the cause and the end point is the effect; for a direct sequential relationship, the starting point comes first and the end point comes later.

3. A fault detection method for a launch vehicle test phase according to claim 2, characterized in that: According to the directed graph, five basic rules are generated: For each directed arc without delay time in the directed graph, the basic rule for generation is that if the value of the graphic code name of the test quantity corresponding to the starting point of the directed arc is equal to the number corresponding to the state of the starting point of the directed arc, then the value of the graphic code name of the test quantity corresponding to the end point of the directed arc is assigned to the number corresponding to the state of the end point of the directed arc; For each directed arc with a delay time in a directed graph, two basic rules are generated, namely: if the value of the graphic code name of the test quantity corresponding to the starting point of the directed arc is equal to the number corresponding to the state of the starting point of the directed arc, a timer is set and starts timing; if the timing result of the timer reaches the delay time of the directed arc, the value of the graphic code name of the test quantity corresponding to the end point of the directed arc is assigned to the number corresponding to the state of the end point of the directed arc; For the "start" state node of the startup instruction quantity in the directed graph, the basic rule generated is that if the test value of the startup instruction quantity is equal to the theoretical value of the startup instruction quantity at the node, the value of the graphical code name of the startup instruction quantity is assigned to the number of the node; For each node corresponding to a switch quantity in the directed graph, two basic rules are generated, which are: if the test value of the switch quantity is equal to the theoretical value of the switch quantity at the node, the value of the graphic code name of the switch quantity is assigned to the number of the node; if the value of the graphic code name of the switch quantity is equal to the number of the node, the expected value of the test value of the switch quantity is assigned to the theoretical value of the switch quantity at the node; For each node corresponding to an analog quantity in the directed graph, the basic rule generated is that if the value of the graphical code name of the analog quantity is equal to the number of the node, the test value of the analog quantity should meet the test value judgment basis of the node; The five generated basic rules are merged into fault detection rules, including: if the antecedent of a basic rule contains the "start" state of the startup instruction quantity of the test item, the basic rule is called a starting point constraint basic rule, and the other basic rules are called non-starting point constraint basic rules; if the antecedent of a non-starting point constraint basic rule is the same as the consequent of a starting point constraint basic rule, the antecedent of the non-starting point constraint basic rule is replaced by the antecedent of the starting point constraint basic rule, and the consequent of the non-starting point constraint basic rule remains unchanged; after processing, if the consequents of two rules are the same, the two rules are merged into one rule: if the antecedents of two rules are both established, the common consequent of the two rules is established.

4. A fault detection method for a launch vehicle test launch phase according to claim 3, characterized in that: According to the fault detection rules for the launch vehicle test launch phase generated in the above steps, the faults in the launch vehicle test launch phase are detected through a rule-based fault diagnosis expert system.

5. A fault detection method for a launch vehicle test launch phase according to claim 2, characterized in that: In the process of constructing the directed graph, a dynamic adaptive adjustment algorithm is also introduced. When unexpected working conditions occur during the test or the external environment changes significantly, the algorithm can automatically adjust the key parameters of the directed graph such as node connection relationship, delay time, etc. according to the preset threshold range and real-time collected data. For example, if the rocket fuel filling test is carried out in cold weather, the low temperature may affect the fuel flow rate and the response characteristics of the pressure sensor. At this time, the algorithm will appropriately extend the directed arc delay time related to the fuel flow rate based on the data feedback from the temperature sensor to ensure that the fault detection rules are still accurately adapted to the actual working conditions, thereby improving the robustness of the system.

6. A fault detection method for a launch vehicle test launch phase according to claim 4, characterized in that: During the operation phase of the rule-based fault diagnosis expert system, a multi-source data fusion verification module is added. This module not only collects test data from the launch vehicle's own systems, but also integrates data from the environment monitoring equipment around the launch site. By cross-validating multi-source data, the risk of misjudgment is further reduced. For example, when an abnormal voltage fluctuation is detected in a certain electrical system of the rocket and a geomagnetic storm occurs near the launch site, the multi-source data fusion verification module will comprehensively analyze the relationship between the two to determine whether the voltage fluctuation is caused by the rocket's own electrical fault or interference from the geomagnetic storm, providing a more comprehensive and accurate basis for fault diagnosis.

Images