A method for automatic health assessment of on-board actuators

Abstract

The present invention belongs to the field of ammunition testing technology, specifically relating to a method for automatically assessing the health of an onboard actuator. The method comprises: collecting actuator commands u and feedback signals y; substituting the collected actuator commands u into a standard model defined according to actuator specifications to simulate and calculate standard feedback ySim; calculating an initial error e0 between the standard feedback and the collected feedback; screening the initial error e0 within the high-dynamic range to obtain a first error e1; low-pass filtering the first error e1 to obtain a second error e2; and determining whether the second error e2 exceeds requirements. This method replaces visual inspection with automated diagnosis, reducing manual workload during mass production, lowering the probability of false positives, improving work efficiency, and lowering production costs. The method, embedded in the onboard computer and executed after the missile is powered on and before launch, can detect faults before launch, thereby improving missile launch and flight reliability.

Description

Technical Field

[0001] The present invention belongs to the technical field of ammunition testing, and in particular relates to a method for automatically evaluating the health of an actuator on a projectile. Background Art

[0002] Moving parts on missiles are more prone to defects or failures than fixed parts. Common moving parts on missiles include the servo and the seeker's position marker. These two moving parts are generally driven by motors. The servo controls the movement of the rudder according to flight control commands, while the position marker adjusts the seeker's direction according to seeker commands. These two components significantly affect the missile's launch and flight reliability, making them essential for inspection during the production process.

[0003] One current technology involves using a data acquisition device to collect the actuator's command and feedback data, plotting them as curves. A human visually inspects the consistency of these two curves. Because command and feedback are dynamic processes that change over time, it's difficult for humans to determine the actuator's consistency at each moment. This requires a fuzzy judgment criterion, which relies on human experience and attention at the moment of verification. This fuzzy judgment is prone to fatigue and makes it difficult to ensure accurate judgments.

[0004] Another current technique involves the onboard computer issuing a standard command to the actuator after power-up, collecting feedback from the actuator, and then determining whether the actuator is functioning properly based on the command and feedback errors within the steady-state operating range. This technique generally only verifies the steady-state error within a typical operating range, making it difficult to cover all operating ranges and verify dynamic characteristics. Summary of the Invention

[0005] (1) Technical issues to be resolved

[0006] The technical problem to be solved by the present invention is: in view of the above-mentioned problems of the existing determination method, how to reduce the inspection workload and increase the inspection coverage.

[0007] (2) Technical solution

[0008] To solve the above technical problems, the present invention provides a method for automatically assessing the health of an actuator on a missile, which includes:

[0009] Step S10: collecting the instruction u and feedback signal y of the actuator;

[0010] Step S20: Substituting the collected actuator instruction u into a standard model set according to the actuator specification to simulate and calculate standard feedback ySim;

[0011] Step S30: Calculate the initial error e0 between the standard feedback and the collected feedback signal, e0=ySim-y;

[0012] Step S40: shielding the high dynamic range, when the actuator instruction change is greater than the specified value, keeping the initial error e0 unchanged for a period of time to obtain the first error e1;

[0013] Step S50: low-pass filtering the first error e1 to obtain a second error e2;

[0014] Step S60: Determine whether the second error e2 exceeds a specified threshold.

[0015] In step S10, the collected instruction u and feedback signal y are further pre-processed, including:

[0016] Step S11: re-sampling the collected feedback signal y of the actuator to obtain feedback at a time point corresponding to the instruction;

[0017] Step S12: selecting the command and feedback signal within the time interval set according to the actuator specification;

[0018] Step S13: Perform median filtering on the command and feedback signals to remove outliers.

[0019] The instruction u is generated by on-board calculation and sent through the RS-422 digital port with a period of 10ms and is received by the acquisition device through the RS-422 port.

[0020] The feedback signal y is collected by the acquisition device as an analog quantity output by the servo and converted into a digital signal with a period of 1ms.

[0021] There is a time difference between the instruction u and the feedback signal y.

[0022] Wherein, in step S11, zero-order hold interpolation is adopted for the feedback y.

[0023] Wherein, in the step S12, the time interval is 0s to 60s.

[0024] Wherein, in step S13, a 3-point sliding window median filter is adopted, and a median is taken for every 3 points.

[0025] In step S20, the standard model is selected as a discrete second-order linear model, and the sampling period is the instruction acquisition period.

[0026] The passband of step S50 is selected to be at least twice the frequency band of the object controlled by the actuator.

[0027] (3) Beneficial effects

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] (1) By replacing visual inspection with automated diagnosis, the workload of people in mass production is reduced, the probability of false detection is reduced, work efficiency is improved, and production costs are reduced.

[0030] (2) This method is embedded in the missile's onboard computer and runs after the missile is powered on and before launch. It can detect faults before launch and improve the missile's launch flight reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flow chart of a specific implementation of the present invention.

[0032] Figure 2 It is a schematic diagram of a state machine for shielding a high dynamic area according to a specific embodiment of the present invention.

[0033] Figure 3 It is a schematic diagram of an error comparison curve for a specific implementation of the present invention. DETAILED DESCRIPTION

[0034] In order to make the purpose, content, and advantages of the present invention more clear, the specific implementation methods of the present invention are further described in detail below with reference to the accompanying drawings and examples.

[0035] To solve the above technical problems, the present invention provides a method for automatically assessing the health of an actuator on a missile, which includes:

[0036] Step S10: collecting the instruction u and feedback signal y of the actuator;

[0037] Step S20: Substituting the collected actuator instruction u into a standard model set according to the actuator specification to simulate and calculate standard feedback ySim;

[0038] Step S30: Calculate the initial error e0 between the standard feedback and the collected feedback signal, e0=ySim-y;

[0039] Step S40: shielding the high dynamic range, when the actuator instruction change is greater than the specified value, keeping the initial error e0 unchanged for a period of time to obtain the first error e1;

[0040] Step S50: low-pass filtering the first error e1 to obtain a second error e2;

[0041] Step S60: Determine whether the second error e2 exceeds a specified threshold.

[0042] In step S10, the collected instruction u and feedback signal y are further pre-processed, including:

[0043] Step S11: re-sampling the collected feedback signal y of the actuator to obtain feedback at a time point corresponding to the instruction;

[0044] Step S12: selecting the command and feedback signal within the time interval set according to the actuator specification;

[0045] Step S13: Perform median filtering on the command and feedback signals to remove outliers.

[0046] The instruction u is generated by on-board calculation and sent through the RS-422 digital port with a period of 10ms and is received by the acquisition device through the RS-422 port.

[0047] The feedback signal y is collected by the acquisition device as an analog quantity output by the servo and converted into a digital signal with a period of 1ms.

[0048] There is a time difference between the instruction u and the feedback signal y.

[0049] Wherein, in step S11, zero-order hold interpolation is adopted for the feedback y.

[0050] Wherein, in the step S12, the time interval is 0s to 60s.

[0051] Wherein, in step S13, a 3-point sliding window median filter is adopted, and a median is taken for every 3 points.

[0052] In step S20, the standard model is selected as a discrete second-order linear model, and the sampling period is the instruction acquisition period.

[0053] The passband of step S50 is selected to be at least twice the frequency band of the object controlled by the actuator.

[0054] Example 1

[0055] Figure 1 This paper presents a specific embodiment of the present invention, which is a method for automatically evaluating a missile servo. During missile assembly, an integration test is performed to inspect the missile components, including the servo. The method for automatically evaluating the servo includes:

[0056] Step S10: Acquire the actuator's command u and feedback signal y, where the command is generated by onboard calculations and sent via the RS-422 digital port with a period of 10ms and received by the acquisition device via the RS-422 port; the feedback is the analog output of the servo collected by the acquisition device and converted into a digital signal with a period of 1ms, with a time difference between the two. Preprocess these two signals, further comprising the following steps:

[0057] Step S11: re-interpolate the feedback y to obtain the feedback at the time point corresponding to the instruction, using zero-order hold interpolation.

[0058] Step S12: selecting the command and feedback signal within a time interval set according to the actuator specification, which is set to 0s to 60s in this embodiment.

[0059] Step S13: Perform median filtering on the command and feedback signals to remove outliers. This embodiment adopts a three-point sliding window median filter, taking a median value for every three points.

[0060] Step S20: Substitute the collected actuator command u into the standard model to simulate and calculate the standard feedback ySim. The standard model of the servo is a second-order linear model, discretized into a difference equation with a 10ms sampling period. Input the command u into the difference equation and iteratively calculate the standard feedback ySim.

[0061] Step S30: Calculate the initial error e0 between the standard feedback ySim and the collected feedback signal y, e0=ySim-y.

[0062] Step S40: shielding high dynamic area, Figure 2 Shows the detailed steps, including two states:

[0063] State S41: Update error. In this state, the first error amount e1 is continuously updated. When the change rate of the instruction u is greater than the specified maximum rate, this state is exited and the state S42 is entered.

[0064] State S42: Maintain the error. In this state, the error is not updated. If the difference between the current time t(k) and the last e1 update time t1(j) is greater than the masking time, the state returns to S41.

[0065] Step S50: low-pass filter the first error e1 to obtain a second error e2. Since the natural frequency of the controlled object in this embodiment is 2 Hz, the filter passband frequency is selected to be 8 Hz, which is 4 times the natural frequency.

[0066] Step S60: Determine whether the error exceeds the requirement.

[0067] Figure 3 The diagram shows the processing result e0 of step S30, the result e1 after shielding the high dynamic area in step S40, and the result e2 after low-pass filtering in step S50. It can be seen that the error of command e0 increases sharply around 25.5s. This error is caused by a delay in the servo responding to the step command, not a fault. Therefore, the error needs to be shielded. e1 is the result of the shielding process, which effectively filters out this effect, but still contains noise. The filtered e2 is smoother.

[0068] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for automatically assessing the health of an actuator on a missile, characterized in that: It includes: Step S10: collecting the instruction u and feedback signal y of the actuator; Step S20: Substituting the collected actuator instruction u into a standard model set according to the actuator specification to simulate and calculate standard feedback ySim; Step S30: Calculate the initial error e0 between the standard feedback and the collected feedback signal, e0=ySim-y; Step S40: shielding the high dynamic range, when the actuator instruction change is greater than the specified value, keeping the initial error e0 unchanged for a period of time to obtain the first error e1; Step S50: low-pass filtering the first error e1 to obtain a second error e2; Step S60: Determine whether the second error e2 exceeds a specified threshold.

2. The method for automatically assessing the health of an actuator on a missile according to claim 1, wherein: In step S10, the collected instruction u and feedback signal y are further pre-processed, including: Step S11: re-sampling the collected feedback signal y of the actuator to obtain feedback at a time point corresponding to the instruction; Step S12: selecting the command and feedback signal within the time interval set according to the actuator specification; Step S13: Perform median filtering on the command and feedback signals to remove outliers.

3. The method for automatically assessing the health of an actuator on a missile according to claim 2, wherein: The instruction u is generated by on-board calculation and sent through the RS-422 digital port with a period of 10ms and is received by the acquisition device through the RS-422 port.

4. The method for automatically assessing the health of an actuator on a missile according to claim 3, wherein: The feedback signal y is collected by the acquisition device as the analog output of the servo and converted into a digital signal with a period of 1ms.

5. The method for automatically assessing the health of an actuator on a missile according to claim 4, wherein: There is a time difference between the instruction u and the feedback signal y.

6. The method for automatically assessing the health of an actuator on a missile according to claim 2, wherein: In step S11, zero-order hold interpolation is performed on the feedback y.

7. The method for automatically assessing the health of an actuator on a missile according to claim 2, wherein: In step S12, the time interval is 0s to 60s.

8. The method for automatically assessing the health of an actuator on a missile according to claim 2, wherein: In step S13, a three-point sliding window median filter is used, and a median is taken for every three points.

9. The method for automatically assessing the health of an actuator on a missile according to claim 1, wherein: In step S20, the standard model is selected as a discrete second-order linear model, and the sampling period is the instruction acquisition period.

10. The method for automatically assessing the health of an actuator on a missile according to claim 1, wherein: The passband in step S50 is selected to be at least twice the frequency band of the object controlled by the actuator.

Images