A method for reducing the false alarm rate of laser ranging based on particle filtering

Abstract

The present invention discloses a method for reducing the false alarm rate of laser ranging based on particle filter, belonging to the technical field of laser ranging false alarm rate, and comprising the following steps: S1: Initialize particle points; S2: Input laser ranging data; S3: Iteratively calculate particle weights; S4: Find the position of the maximum weight; S5: Calculate local particle filter and output the result; S6: Update or not update the particle neighborhood width; S7: Linearly extrapolate to calculate the estimated value at the next moment; S8: Return to S2 and start the loop again. The method for reducing the false alarm rate of laser ranging based on particle filter proposed by the present invention is based on particle filter, takes the maximum value position of particles in the state space as the filtering center, calculates the neighborhood width of particles through residual iteration, reduces the false alarm rate of laser ranging, and finally estimates the laser ranging state at the next moment through linear extrapolation, effectively reducing the appearance of false alarm targets, making the laser ranging data stable and with rapid response.

Description

Technical Field

[0001] The present invention relates to the technical field of laser ranging false alarm rate, and particularly relates to a method for reducing the laser ranging false alarm rate based on particle filtering. Background Art

[0002] In a laser ranging and tracking system, first, a photoelectric device is used to detect a target, and then the laser ranging is controlled to measure the distance to the target to achieve three-dimensional positioning of the target and draw the three-dimensional trajectory of the target. To achieve stable drawing of the target three-dimensional trajectory, it is required that the laser ranging responds quickly, and the data is stable and accurate. Quick response requires a high frequency of the laser rangefinder and a fast algorithm. The accuracy is ensured by the design of the laser rangefinder itself. However, in terms of stability, when measuring the distance to a long-distance target, the laser has high sensitivity and is easily interfered by other targets. When positioning a near-low-altitude target, due to the complex use environment, the problem of high false alarm rate of ranging will occur.

[0003] If the initial state of the known target is inaccurate, the initial state value of the particle is too far from the true value, resulting in a small number of particles that meet the requirements in the initial particle swarm. Therefore, resampling will continuously occur in the initial stage of filtering, the number of effective particles will suddenly decrease, resulting in poor filtering effect, and even divergence may occur.

[0004] To solve the problem of high false alarm rate of laser ranging, a method for reducing the laser ranging false alarm rate based on particle filtering is designed and implemented on an FPGA, which can effectively reduce the occurrence of false alarm targets, make the laser ranging data stable, and respond quickly. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for reducing the laser ranging false alarm rate based on particle filtering, taking the maximum value position of the particle in the state space as the filtering center, calculating the neighborhood width of the particle through residual iteration to reduce the false alarm rate of laser ranging, and finally estimating the laser ranging state at the next moment through linear extrapolation, effectively reducing the occurrence of false alarm targets, making the laser ranging data stable, and responding quickly, so as to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: A method for reducing the laser ranging false alarm rate based on particle filtering, including the following steps:

[0007] S1: Initialize particle points;

[0008] S2: Input laser ranging data;

[0009] S3: Iteratively calculate particle weights;

[0010] S4: Find the position of the maximum weight;

[0011] S5: Calculate the local particle filter and output the result;

[0012] S6: Update the particle neighborhood width or not;

[0013] S7: Calculate the estimated value at the next moment by linear extrapolation;

[0014] S8: Return to S2 and start the loop again.

[0015] Further, for S1, the initialization of particles can call the particle system module. By assuming that the initial state of the target is known, particles are selected, and each individual particle performs operations. Set the random attributes of particle generation, including acceleration, particle transparency, particle color, and particle rotation angle.

[0016] Further, the initial value of the particle sets the acceleration of particle movement, which is the acceleration before reaching the defined speed. A positive value on the x-axis or y-axis will cause the particle to move upward and to the right, while a negative value will cause the particle to move downward and to the left;

[0017] Particle rotation is generated and allows defining the minimum / maximum values of the rotation angle.

[0018] Further, for S2, input the distance L from the particle emission end to the particle beam receiving end of the particle beam transmitter, and input the time interval t from the particle beam emission moment to the receiving moment.

[0019] Further, for S3, the iterative calculation of particle weights includes the following steps:

[0020] S31: Calculate the current particle system observation value from the particle state transition matrix or equation;

[0021] S32: Calculate the observation value from the current state measurement state matrix or equation of the particle;

[0022] S33: Calculate the function value according to the current observation value of the particle in S2 and the observation value of the system to obtain the particle weight value.

[0023] Further, for S4, normalize the weights of the particles, sample all particles to obtain the maximum value greater than the particle weight.

[0024] Further, for S5, the particle filter algorithm structure is adopted. Multiple targets and corresponding particle groups are formed by particle clustering. The target particle groups are detected and the tracking target particle groups are fused. By extracting high-quality particles from the two target particle groups and using crossover operations to obtain new particles, the obtained target tracking particle group contains high-quality particles from the two particle groups. The particle filter algorithm will eliminate particles with low weights and let particles with high weights generate more particles. The algorithm converges towards the place with high weights, and the maximum value position of the particles in the state space is used as the filtering center.

[0025] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0026] A method for reducing the false alarm rate of laser ranging based on particle filtering proposed by the present invention forms an initial particle swarm by selecting multiple initial values as the center, determines the strength of its rationality according to the number of remaining effective particles emitted from each particle swarm to the receiving end, and selects the particle swarm with the most effective particle data as the initial particle; based on particle filtering, the maximum value position of the particle in the state space is used as the filtering center, and the neighborhood width of the particle is calculated by residual iteration to reduce the false alarm rate of laser ranging. Finally, the state of laser ranging at the next moment is estimated by linear extrapolation, effectively reducing the occurrence of false alarm targets, making the laser ranging data stable and with a rapid response. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is the overall flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying 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 of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0029] Please refer to Figure 1 , a method for reducing the false alarm rate of laser ranging based on particle filtering, includes the following steps:

[0030] Step 1: Initialize the particle points. The initialization of the particles can adjust the particle system module. By assuming that the initial state of the target is known, particles are selected, and each individual particle performs operations. Set the random attributes of particle generation, including acceleration, particle transparency, particle color, and particle rotation angle;

[0031] The initial value of the particle sets the acceleration of the particle's movement, which is the acceleration before reaching the defined speed. A positive value on the x-axis or y-axis will cause the particle to move upward and to the right, while a negative value will cause the particle to move downward and to the left; particle rotation is generated and allows the definition of the minimum / maximum values of the rotation angle;

[0032] By selecting multiple initial values as the center, an initial particle swarm is formed. The strength of its rationality is determined according to the number of remaining effective particles emitted from each particle swarm to the receiving end, and the particle swarm with the most effective particle data is selected as the initial particle;

[0033] Step 2: Input of laser ranging data. Input the distance L from the particle emission end to the particle beam receiving end of the particle beam transmitter, and input the time interval t from the particle beam emission time to the reception time;

[0034] Step 3: Iterative calculation of particle weights;

[0035] Step 1: Calculate the current particle system observation value from the particle state transition matrix or equation;

[0036] Step 2: Calculate the observation value from the current state measurement state matrix or equation of the particle;

[0037] Step 3: Calculate the function value according to the current observation value of the particle in Step 2 and the observation value of the system to obtain the particle weight value;

[0038] Step 4: Find the position of the maximum weight, normalize the weights of the particles, and sample all particles to obtain the maximum value greater than the particle weight value;

[0039] Step 5: Calculate the local particle filter and output the result. Adopt the particle filter algorithm structure, form multiple targets and corresponding particle groups through particle clustering, detect and track the target particle groups for fusion, extract high-quality particles from the two target particle groups, use crossover operations to obtain new particles, and the obtained target tracking particle group contains high-quality particles from the two particle groups. The particle filter algorithm will eliminate particles with low weights and let particles with high weights generate more particles. The algorithm converges towards the place with high weights, and takes the maximum value position of the particles in the state space as the filter center;

[0040] Step 6: Update or not update the particle neighborhood width;

[0041] Step 7: Calculate the estimated value at the next moment by linear extrapolation;

[0042] Step 8: Return to Step 2 and start the loop again.

[0043] Based on particle filtering, taking the maximum value position of the particles in the state space as the filter center, calculate the particle filter state output of the local particles within the neighborhood width as the laser rangefinder state output, and calculate the neighborhood width of the particles through residual iteration to reduce the false alarm rate of laser ranging. Finally, estimate the laser ranging state at the next moment through linear extrapolation. The method proposed by the invention can effectively reduce the appearance of false alarm targets, make the laser ranging data stable and respond quickly. This method is implemented on FPGA.

[0044] In summary, the method for reducing the false alarm rate of laser ranging based on particle filtering in the present invention forms an initial particle swarm by selecting multiple initial values as the center, determines the strength of its rationality according to the number of remaining effective particles emitted from each particle swarm to the receiving end, and selects the particle swarm with the largest number of effective particle data as the initial particle; based on particle filtering, the maximum value position of the particle in the state space is used as the filtering center, and the neighborhood width of the particle is calculated by residual iteration to reduce the false alarm rate of laser ranging. Finally, the laser ranging state at the next moment is estimated by linear extrapolation, effectively reducing the occurrence of false alarm targets, making the laser ranging data stable and the response rapid.

[0045] As described above, the above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes should be covered within the protection scope of the present invention.

Claims

1. A method for reducing the false alarm rate of laser ranging based on particle filtering, characterized in that It includes the following steps: S1: Initialize particle points; S2: Input laser ranging data; S3: Iteratively calculate particle weights, including the following steps: S31: Calculate the current particle system observation value from the particle state transition matrix or equation; S32: Calculate the observation value from the current state measurement state matrix or equation of the particle; S33: Calculate the function value based on the current observation value of the particle in S2 and the observation value of the system to obtain the particle weight value; S4: Find the position of the maximum weight; S5: Calculate local particle filtering and output the result; form multiple targets and corresponding particle groups through particle clustering, detect and track the target particle groups for fusion, extract high-quality particles from the two target particle groups, use crossover operations to obtain new particles, and the resulting target tracking particle group contains high-quality particles from both particle groups. The particle filtering algorithm will eliminate particles with low weights and let particles with high weights generate more particles. The algorithm converges towards the place with high weights, and takes the maximum value position of the particles in the state space as the filtering center; S6: Update or not update the particle neighborhood width; S7: Linearly extrapolate to calculate the estimated value at the next moment; S8: Return to S2 and start the loop again.

2. The method for reducing the false alarm rate of laser ranging based on particle filtering according to claim 1, wherein For S1, the initialization of particles can call the particle system module. Select particles under the assumption that the initial state of the target is known, and each individual particle performs operations. Set the random attributes of particle generation, including acceleration, particle transparency, particle color, and particle rotation angle.

3. The method for reducing the false alarm rate of laser ranging based on particle filtering according to claim 2, characterized in that, The initial value of the particle sets the acceleration of particle movement, which is the acceleration before reaching the defined speed. A positive value on the x-axis or y-axis will cause the particle to move upward and to the right, while a negative value will cause the particle to move downward and to the left; Particle rotation is generated and allows defining the minimum / maximum values of the rotation angle.

4. The method for reducing the false alarm rate of laser ranging based on particle filtering according to claim 1, wherein, For S2, input the distance L from the particle emission end of the particle beam transmitter to the particle beam receiver, and input the time interval t from the particle beam emission moment to the reception moment.

5. A method for reducing the false alarm rate of laser ranging based on particle filtering according to claim 1, characterized in that For S4, normalize the weights of the particles and sample all particles to obtain the maximum value greater than the particle weight.

Images