Multi-source information fusion signal feature extraction method of penetration fuze recognition warhead target hitting process

Abstract

The invention discloses a multi-source information fusion signal feature extraction method for recognizing a warhead target hitting process by a penetration fuze, and belongs to the technical field of penetration fuzes. An identification mode of multi-information fusion of an acceleration sensor and an inertia switch is adopted, and overload signals are distinguished from electromagnetic interference signals under strong coupling and strong impact environments in a penetration endpoint detection mode of an energy-entropy ratio in a warhead targeting process; meanwhile, the problems that the impact overload signal of a penetration warhead on a soft medium target is small, and the zero point starting is unreliable and the anti-electromagnetic interference capability is weak under the condition that a single high-g acceleration sensor is adopted by a fuze to recognize a low overload condition and a projectile body collides with a weak target are solved by combining inertia switch signal combination; and the fuse can be reliably started to ignite when the projectile body penetrates a weak target.

Description

Technical Field

[0001] The invention belongs to the technical field of penetration fuzes, and in particular relates to a signal feature extraction method for a penetration fuze identifying a warhead target impact process by multi-source information fusion. Background Art

[0002] The penetration fuze is the key to the efficient destruction of hard targets by penetrating ammunition, and its core is the adaptive detonation control technology under high overload conditions. At present, with the rapid development of navigation systems and electronic information technology, the precision guidance technology of high-precision weapons is becoming more and more mature. Precision-guided weapons are equipped with a front cabin at the front end of the warhead. The front cabin structure contains a seeker, inertial navigation and flight control devices, which are interconnected with satellites, Beidou system and other on-board systems. The penetration fuze follows the large aspect ratio penetrating warhead to strike high-value enemy targets. The penetration process needs to withstand tens of thousands of g of high overload and strong impact. In order to ensure that the penetration fuze can function normally during the high overload and strong impact process, the penetration fuze is generally installed at the rear end of the penetration warhead. After the penetrating warhead penetrates the target, the fuze identifies the overload signal transmitted by the projectile through the acceleration sensor as the input signal for the target identification of the penetration fuze. When the large aspect ratio penetrating warhead with a front cabin penetrates a weak target at high speed, the front cabin of the warhead contacts and squeezes the target. The front cabin structure is not completely destroyed and becomes a buffer structure between the oval part of the warhead and the target. The impact energy felt by the front cabin is transmitted to the overload signal felt by the fuze at the tail of the projectile, which will be further attenuated. The peak value of the overload signal is small (between 500g and 1000g) and the pulse width is narrow. It is difficult to determine the threshold value using the traditional threshold recognition method. After the projectile hits the target, the threshold of the overload signal is too high and the ignition cannot be reliably started. If the trigger threshold is set too low, there is a problem of weak anti-electromagnetic interference capability. Summary of the invention

[0003] Aiming at the problems that the penetration fuze cannot be reliably started and fired when a large aspect ratio warhead penetrates a weak target at high speed and has weak anti-electromagnetic interference capability, the present invention proposes a signal feature extraction method for a penetration fuze identifying a warhead impact process by multi-source information fusion, adopts an identification mode of multi-information fusion of an acceleration sensor + an inertial switch, and uses an energy entropy ratio penetration endpoint detection method to distinguish the overload signal of the warhead impact process from the electromagnetic interference signal under a strong coupling and strong impact environment. At the same time, combined with the inertial switch signal (the inertial switch is not turned on below 100g and is reliably turned on above 200g), the problem of small overload signal of the penetration warhead impacting a soft medium target, unreliable zero-point start-up and weak anti-electromagnetic interference capability of the projectile body when identifying a low overload condition and a weak target is solved by the fuze using a single high-g acceleration sensor.

[0004] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is as follows:

[0005] The present invention proposes a signal feature extraction method for a multi-source information fusion penetration fuze identification warhead target process, comprising the following steps:

[0006] In the first step, the overload signal a(n) is multiplied by a Hanning window function with a finite length of N to perform frame division, and the inertial switch signal is synchronously collected. When the fuze detonation control module recognizes that the inertial switch signal level is high, the switch counter value cnt is increased by 1, and the next step is executed. When the fuze detonation control module recognizes that the inertial switch signal level is low, the second step is directly executed.

[0007] y i (n) = a(n)·w(in)

[0008] In the formula, y i a(n) is the overload signal of the i-th frame, w(n) is the Hanning window function, i is the windowing position in the overload signal a(n), and w(in) is not 0 only when n is between i-N+1 and i; a(n) is the overload signal at the current moment.

[0009] In the second step, after the overload signal is windowed and framed, the energy of each frame is calculated.

[0010]

[0011] In the formula, E i is the overload signal y of the i-th frame i (n) energy, a is a constant. When a takes a larger value, it can smooth out the drastic change in energy. Therefore, proper selection of a can help distinguish overload signals from electromagnetic interference signals.

[0012] The third step is to perform Fourier transform on each frame of overload signal to obtain its spectrum Y i (k), determine the normalized spectral probability density of each frequency component of the overload signal of each frame:

[0013]

[0014] In the formula, p i (k) is the overload signal y of the i-th frame i The FFT spectrum Y of (n) i (k) Normalized probability density corresponding to the kth frequency component, M is the FFT length.

[0015] Based on this, the short-time spectral entropy of the overload signal of each frame is calculated:

[0016]

[0017] In the formula, H i is the overload signal y of the i-th frame i(n) short-time spectral entropy;

[0018] The fourth step is to use the calculated energy E of the overload signal per frame i and short-time spectral entropy H i , calculate the energy entropy ratio of the overload signal of each frame,

[0019]

[0020] In the formula, R i is the overload signal y of the i-th frame i (n) energy entropy ratio, energy per frame E i Calculated from step 2, the spectral entropy H i Calculated from step 3;

[0021] The fifth step is to identify that the switch counter value cnt ≥ m times, and the energy entropy ratio of the overload signal Ri ≥ K, confirm that the target collision identification is valid, the fuse is fired and started, and the process ends, otherwise it returns to the first step.

[0022] In the present invention, the value of m is greater than or equal to twice, and K is usually between 0.5 and 1.

[0023] Compared with the prior art, the signal feature extraction method of the multi-source information fusion penetration fuze identification warhead target process of the present invention has the following beneficial effects:

[0024] (1) The present invention adopts an identification mode of combined identification of an overload sensor and an inertial switch, and uses the high sensitivity of the inertial switch under low overload conditions to make up for the problem that the resolution of a single high-g acceleration sensor is insufficient under low overload conditions, resulting in the fuze being unable to reliably start and fire after the projectile hits the target;

[0025] (2) The overload signal of the present invention uses an energy entropy ratio-based penetration endpoint detection to distinguish between overload signals and interference signals in the process of hitting the target, so as to solve the problem that the threshold setting method of the penetration fuze cannot effectively distinguish the characteristics of the interference signal and the overload signal of hitting the target, and the fuze mistakenly identifies that the projectile hits the target and starts firing prematurely due to the abnormal electromagnetic interference signal. The energy entropy ratio coefficient after overload signal processing is combined with the output signal of the inertial switch to confirm the warhead hitting the target, thereby improving the reliability and accuracy of the penetration fuze in the field of weak target identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is the overload signal, inertial switch signal and overload signal energy entropy ratio of 810m / s 1200kg warhead (warhead length (2.2m) with front cabin 1.4m) penetrating 10mm steel plate (907A). DETAILED DESCRIPTION

[0027] In one embodiment of the present invention, a 2.2m long projectile penetrates a 10mm layer 907A steel plate at an initial speed of 810m / s, and the fuze is required to be reliably activated after identifying the intruding target. The overload signal collected by the acceleration sensor at the fuze part and the inertial switch signal are as follows: Figure 1 As shown, Figure 1 The top curve is the overload signal. Figure 1 The middle curve is the inertia switch signal. Figure 1 The bottom curve is the overload signal energy entropy ratio.

[0028] The present invention provides a specific implementation method of a signal feature extraction method for a penetration fuze identification warhead target process based on multi-source information fusion, which specifically includes the following steps:

[0029] Step 1: The overload signal a(n) is multiplied by the Hanning window function with a window length N of 255 for frame division, and the inertial switch signal is synchronously collected. When the fuze detonation control module recognizes that the inertial switch signal level is high, the switch counter value cnt is increased by 1, and the next step is executed. When the fuze detonation control module recognizes that the inertial switch signal level is low, the second step is directly executed;

[0030] y i (n) = a(n)·w(in)

[0031] In the formula, y i (n) is the overload signal of the i-th frame, w(n) is the Hanning window function, i is the windowing position in the overload signal a(n), and w(in) is not 0 only when n is between i-N+1 and i;

[0032] Step 2: After the overload signal is windowed and framed, the energy of each frame is calculated:

[0033]

[0034] In the formula, E i is the overload signal y of the i-th frame i (n) energy, a is a constant. When a takes a larger value, it can smooth out the drastic change in energy. Therefore, properly selecting a can help distinguish overload signals from interference signals. In this case, the value of a is 2.

[0035] Step 3, perform Fourier transform on each frame of overload signal to obtain its spectrum, and determine the normalized spectrum probability density function of each frequency component of each frame signal:

[0036]

[0037] In the formula, p i (k) is the overload signal y of the i-th frame i The FFT spectrum Y of (n)i (k) Normalized probability density corresponding to the kth frequency component, M is the FFT length.

[0038] Based on this, the short-time spectral entropy of the overload signal of each frame is calculated:

[0039]

[0040] In the formula, H i is the overload signal y of the i-th frame i The short-time spectral entropy of (n):

[0041] Step 4: Use the energy E per frame calculated using the above steps i and spectral entropy H i , calculate the energy entropy ratio of each frame signal:

[0042]

[0043] In the formula, R i is the overload signal y of the i-th frame i (n) energy entropy ratio, energy per frame E i Calculated from step 2, the spectral entropy H i Calculated from step 3;

[0044] Step 5, if the switch counter value cnt≥3 times and the overload signal entropy ratio coefficient Ri≥0.8 is identified, the target identification is confirmed to be valid, the fuse is fired and the process ends, otherwise it returns to step 1.

Claims

1. A signal feature extraction method for a penetration fuze identifying a warhead during its impact process based on multi-source information fusion, characterized in that: The Method The following steps are involved: In the first step, the overload signal a(n) is multiplied by a Hanning window function with a finite length of N to perform frame division, and the inertial switch signal is synchronously collected. When the fuze detonation control module recognizes that the inertial signal level is high, the switch counter value cnt is increased by 1, and the next step is executed. When the fuze detonation control module recognizes that the inertial signal level is low, the second step is directly executed. In the second step, after the overload signal is windowed and framed, the energy of each frame is calculated; The third step is to perform Fourier transform on each frame of overload signal to obtain its spectrum Y i (k), determine the normalized spectral probability density p of each frequency component of the overload signal of each frame i (k), calculate the short-time spectral entropy of the overload signal of each frame: In the formula, H i is the overload signal y of the i-th frame i (n) short-time spectral entropy; The fourth step is to use the calculated energy E of the overload signal per frame i and short-time spectral entropy H i , calculate the energy entropy ratio of the overload signal of each frame, In the formula, R i is the overload signal y of the i-th frame i (n) energy-entropy ratio; The fifth step is to identify that the switch counter value cnt ≥ m times, and the energy entropy ratio of the overload signal Ri ≥ K, confirm that the target collision identification is valid, the fuse is fired and started, and the process ends, otherwise it returns to the first step.

2. The signal feature extraction method for the process of identifying the warhead of a penetration fuze by multi-source information fusion according to claim 1 is characterized by: The first step specifically includes: multiplying the overload signal a(n) by a Hanning window function of a finite length N for framing; y i (n)=a(n)·w(i-n) In the formula, y i (n) is the overload signal of the i-th frame, w(n) is the Hanning window function, i is the windowing position in the overload signal a(n), and w(in) is not 0 only when n is between i-N+1 and i.

3. The signal feature extraction method for the process of identifying the warhead of a penetration fuze by multi-source information fusion according to claim 1 is characterized in that: The second step specifically includes: Calculate the energy of each frame, In the formula, E i is the overload signal y of the i-th frame i (n), a is a constant.

4. The signal feature extraction method for the process of identifying the warhead of a penetration fuze by multi-source information fusion as claimed in claim 3 is characterized by: When a takes a larger value, it can smooth out drastic changes in energy. The value of a is 2.

5. The signal feature extraction method for the process of identifying the warhead of a penetration fuze hitting a target by multi-source information fusion as claimed in claim 1 is characterized by: The third step specifically includes: determining the normalized spectral probability density of each frequency component of each frame of the overload signal: In the formula, p i (k) is the overload signal y of the i-th frame i The FFT spectrum Y of (n) i (k) Normalized probability density corresponding to the kth frequency component, M is the FFT length.

6. The signal feature extraction method for the process of identifying the warhead of a penetration fuze by multi-source information fusion according to claim 1 is characterized in that: In the fifth step, the value of m is greater than or equal to 2 times.

7. The signal feature extraction method for the process of identifying the warhead of a penetration fuze by multi-source information fusion according to claim 1 is characterized by: In the fifth step, K is usually between 0.5 and 1.

8. The signal feature extraction method for the process of identifying the warhead of a penetration fuze by multi-source information fusion according to claim 1 is characterized in that: The inertia switch in the penetration fuze does not conduct below 100g, but conducts reliably above 200g.

Images