A dimensionality reduction method for performance calculation of phased array radar system simulation models

Abstract

This invention discloses a method for dimensionality reduction in performance calculation of phased array radar system simulation models, relating to the field of radar signal processing technology. The method includes the following steps: extracting multiple sets of input parameter samples for a phased array radar system within a given interval; sequentially substituting the input parameter samples into the phased array radar system simulation model to calculate the radar system performance response; changing only one parameter dimension's sample data each time, and calculating the influencing parameters corresponding to its performance response through the phased array radar system simulation model; when all parameter dimensions have been calculated, obtaining all influencing parameters for each radar system performance, removing parameters with less influence to reduce the parameter dimensionality required for model calculation; substituting the influencing parameters into the phased array radar system simulation model to calculate the corresponding phased array radar system performance response; and obtaining the order of importance of parameters affecting the phased array radar system performance, providing a reference for improving radar performance calculation efficiency.

Description

Technical Field

[0001] This invention generally relates to the field of radar signal processing technology, and specifically to a method for performance calculation dimensionality reduction of a phased array radar system simulation model. Background Technology

[0002] Radar systems, due to their superior target parameter (such as angle and range) estimation and detection capabilities, are widely used in military and defense fields (military intelligence reconnaissance, early warning and guidance, weapons control, etc.) and civilian fields (traffic control, weather forecasting, resource exploration, etc.). With the widespread application of radar, radar detection simulation technology can simulate radar detection systems to perform detection operations, providing an economical and effective solution for radar detection system design, parameter optimization, and performance evaluation.

[0003] Existing phased array radars (PARs) have fundamentally solved the inherent problems of traditional mechanically scanned radars. Under the same aperture and operating wavelength, phased array radars are superior to traditional radars in terms of reaction speed, target update rate, multi-target tracking capability, resolution, multi-functionality, and electronic countermeasures capability. However, compared to traditional mechanically scanned radars, phased array radar systems are more complex, are mostly used in the military field, have high technical requirements, involve more parameters, have strong coupling between parameters, and have more stringent performance calculation time, which leads to inaccurate detection results, which in turn affects decision-making timing and produces a series of consequences.

[0004] Since clutter objectively exists and cannot be completely eliminated, the parameters required for performance calculation will fluctuate. Therefore, we introduce the concept of uncertainty and treat the parameters as interval parameters to eliminate the influence caused by fluctuations. We propose a dimensionality reduction method for performance calculation of phased array radar system simulation models to solve the above problems. Summary of the Invention

[0005] In view of the above-mentioned defects or deficiencies in the existing technology, it is desirable to provide a performance calculation dimensionality reduction method that reduces the dimensionality of performance parameters of phased array radar systems, simplifies calculation steps, and improves calculation efficiency.

[0006] This invention provides a method for dimensionality reduction in performance calculation of a phased array radar system simulation model, comprising the following steps:

[0007] Multiple sets of input parameter samples from the phased array radar system are collected; each set of input parameter samples includes multiple parameter dimensions.

[0008] Each set of input parameter samples is substituted into the phased array radar system simulation model in turn to calculate the corresponding radar system performance response.

[0009] Each time, only the sample data of one of the parameter dimensions is changed, and the other parameter dimensions are set to the midpoint of the interval. The corresponding radar system performance response is calculated by the phased array radar system simulation model.

[0010] When all the parameter dimensions have been calculated once, all the parameters affecting the performance of each radar system are obtained; wherein, when the radar system performance responses obtained from multiple sample data of the same parameter dimension are inconsistent, that parameter dimension is the influencing parameter.

[0011] Substitute the samples of the influencing parameters into the simulation model of the phased array radar system respectively to calculate the corresponding performance response of the phased array radar system;

[0012] Normalization processing is performed on multiple sets of input parameter samples and the corresponding performance responses of the phased array radar system.

[0013] Using the normalized influence parameters as input variables and the corresponding normalized phased array radar system performance response sequence as output response, a polynomial response surface proxy model is established.

[0014] If the polynomial response surface surrogate model satisfies the preset index, calculate the ratio of the variance of each input parameter sample to the total variance to obtain the sensitivity index of different input parameter samples.

[0015] Based on the aforementioned sensitivity index, the order of importance of parameters affecting the performance of the phased array radar system is determined.

[0016] According to the technical solution provided by the present invention, multiple sets of input parameter samples of a multi-output radar system are adopted, specifically including the following steps:

[0017] S1: Acquire multiple initial input parameters of the phased array radar system; each initial input parameter has a corresponding data range.

[0018] According to the technical solution provided by the present invention, after establishing the polynomial response surface surrogate model, before calculating the ratio of the variance of each input parameter sample to the total variance, the following steps are also included:

[0019] If the polynomial response surface proxy model does not meet the preset index, then execute step S1 above.

[0020] According to the technical solution provided by the present invention, obtaining the order of importance of parameters affecting the performance of the phased array radar system specifically includes the following steps:

[0021] The parameters affecting the performance of the radar system are arranged in descending order of their sensitivity indices.

[0022] According to the technical solution provided by the present invention, the performance response of the radar system includes: maximum detection range, self-defense jamming range, detection probability, range accuracy, azimuth accuracy, and elevation accuracy.

[0023] According to the technical solution provided by the present invention, the initial input parameters include: signal / interference frequency, signal bandwidth, pulse width, pulse accumulation number, radar transmit power, transmit gain, receive gain, overall loss, azimuth beamwidth, elevation beamwidth, angle measurement error slope, interference bandwidth, interference power, and interference transmit gain.

[0024] According to the technical solution provided by the present invention, multiple sets of input parameter samples of a phased array radar system are extracted according to the following steps:

[0025] Multiple sets of input parameter samples for the phased array radar system were extracted using the Latin hypercube sampling method.

[0026] In summary, this invention specifically discloses a detailed process for a performance calculation dimensionality reduction method for a phased array radar system simulation model. This invention involves taking multiple sets of input parameter samples from the phased array radar system; each set of input parameter samples includes multiple parameter dimensions; each set of input parameter samples is sequentially substituted into the phased array radar system simulation model to calculate the corresponding radar system performance response; each time, only the sample data of one parameter dimension is changed, while the remaining parameter dimensions are set to the midpoint of the interval, and the radar system performance response is calculated through the phased array radar system simulation model until all parameter dimensions have been calculated once, obtaining all the influencing parameters of each radar system performance; and then, all the influencing parameters corresponding to the performance response of each radar system are substituted into the phased array radar system simulation model. The system simulation model is used to calculate the corresponding phased array radar system performance response. Multiple sets of input parameter samples and their corresponding phased array radar system performance responses are normalized. Using the normalized influencing parameters as input variables and the corresponding normalized phased array radar system performance response sequence as the output response, a polynomial response surface surrogate model is established. If the polynomial response surface surrogate model meets preset indices, the ratio of the variance of each input parameter sample to the total variance is calculated to obtain sensitivity indices for different input parameter samples. Based on these sensitivity indices, the order of importance of parameters affecting the phased array radar system performance is obtained.

[0027] Compared to existing phased array radar models, which suffer from numerous parameters and strong coupling between parameters, leading to low computational efficiency, this invention uses local sensitivity analysis of the phased array radar system to obtain the influencing parameters and dimensions of different performance characteristics. By classifying and organizing these parameters, the dimensionality of the performance parameters of the phased array radar system can be effectively reduced while improving computational efficiency. Attached Figure Description

[0028] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0029] Figure 1 A flowchart of a dimensionality reduction method for performance calculation of a phased array radar system simulation model.

[0030] Figure 2 A flowchart of the overall process for dimensionality reduction methods to consider the performance calculation of phased array radar system simulation models.

[0031] Figure 3 This is a line graph illustrating the performance response of the input parameter sequence corresponding to the maximum detection distance as a function of the input parameters.

[0032] Figure 4 This is a piecewise linear diagram illustrating the performance response of the input parameter sequence corresponding to the self-defense interference range as a function of the input parameters.

[0033] Figure 5 This is a piecewise linear diagram illustrating the performance response of the input parameter sequence corresponding to the detection probability as a function of the input parameters.

[0034] Figure 6 This is a piecewise linear diagram illustrating the performance response of the input parameter sequence corresponding to the distance accuracy as a function of the input parameters.

[0035] Figure 7 This is a line graph illustrating the performance response of the input parameter sequence corresponding to the azimuth accuracy as a function of the input parameters.

[0036] Figure 8 This is a line graph illustrating the performance response of the input parameter sequence corresponding to pitch angle accuracy as a function of the input parameters. Detailed Implementation

[0037] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0038] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0039] Example 1

[0040] Please refer to Figure 1 and Figure 2 The flowchart shown is a first embodiment of the performance calculation dimensionality reduction method for a phased array radar system simulation model provided by the present invention, which includes the following steps:

[0041] S10. Take multiple sets of input parameter samples from the phased array radar system; each set of input parameter samples includes multiple parameter dimensions;

[0042] The process of extracting multiple sets of input parameter samples from a phased array radar system includes the following steps:

[0043] Step S1: Obtain multiple initial input parameters of the phased array radar system; each initial input parameter has a corresponding data range;

[0044] The initial input parameters include: signal / jamming frequency, signal bandwidth, pulse width, pulse accumulation number, radar transmit power, transmit gain, receive gain, overall loss, azimuth beamwidth, elevation beamwidth, angle measurement error slope, jamming bandwidth, jamming power, and jamming transmit gain.

[0045] Here, the data range corresponding to the initial input parameters, for example, the initial data range for the signal / interference frequency is [3×10]. 8 1×10 9 ];

[0046] Furthermore, the Latin hypercube sampling method can be used to extract multiple sets of input parameter samples for the phased array radar system.

[0047] S20. Substitute the input parameters from each set of input parameter samples into the phased array radar system simulation model in turn, and calculate the corresponding radar system performance response.

[0048] The performance response of a radar system includes: maximum detection range, self-defense jamming range, detection probability, range accuracy, azimuth accuracy, and elevation accuracy.

[0049] Furthermore, the performance response and input parameters of the phased array radar system are shown in Table 1:

[0050] Table 1. Phased Array Radar System Performance and Input Parameters

[0051]

[0052] S30. Each time, only the sample data of one parameter dimension is changed, and the samples of the remaining parameters are all set to the midpoint of the interval. The performance response of the radar system is calculated by the phased array radar system simulation model.

[0053] S40. When all parameter dimensions have been calculated once, all influencing parameters of the performance response of each radar system are obtained; wherein, when the radar system performance response obtained from multiple sample data of the same parameter dimension is inconsistent, that parameter dimension is the influencing parameter.

[0054] like Figures 3 to 8As shown, these are all the influencing parameters corresponding to the performance response of each radar system; that is, the sensitivity of a parameter dimension is measured by the response change caused by a single parameter dimension; and it is determined whether there is an impact on the phased array radar system response. If there is an impact, it is used as an influencing parameter of the phased array radar system.

[0055] S50. Substitute the samples of the influencing parameters into the phased array radar system simulation model respectively, and calculate the corresponding phased array radar system performance response.

[0056] The relationship between the performance response of the phased array radar system and the influencing parameters is shown in Table 2.

[0057] Table 2. Influencing parameters for each performance response of the phased array radar system.

[0058]

[0059] S60. Normalize the set of multiple input parameter samples and the corresponding performance response set of the phased array radar system.

[0060] The initial interval corresponding to all input parameter samples and the corresponding radar system performance response set is transformed to the interval [0, 1] to further reduce the interval range and eliminate the influence of the order of magnitude of each parameter on the performance analysis of the phased array radar system, so as to perform integration operations.

[0061] S70. Using the normalized influence parameters as input variables and the corresponding and normalized phased array radar system performance response as output response, a polynomial response surface proxy model is established.

[0062] Each output performance corresponds to a polynomial response surface surrogate model; for each model, integration is performed in the interval [0, 1] to obtain the variance of each input parameter and the total variance of each model; it is then determined whether it meets the preset index, for example, the index is set to an error value of 1%, when the error value is less than or equal to 1%, the requirement is met, otherwise the requirement is not met.

[0063] The polynomial response surface surrogate model refers to a method that assumes the input-output relationship of the model is a polynomial. After determining the order of the polynomial, the coefficients of each term are obtained through the least squares method. Once the response surface is constructed, the influence of each variable on the output can be easily derived from the coefficients. For high-dimensional problems, the response surface method can effectively eliminate insensitive variables, reducing the optimization difficulty. It can still converge quickly for models with errors, playing an important role in this invention.

[0064] S80. If the polynomial response surface surrogate model meets the preset index, calculate the ratio of the variance of each input parameter to the total variance to obtain the sensitivity index of different input parameters.

[0065] If the polynomial response surface proxy model does not meet the preset index, then step S1 above is executed to re-divide the sub-intervals until the model accuracy meets the requirements.

[0066] S90. Based on the sensitivity indicators, obtain the order of importance of the parameters affecting the performance of the phased array radar system.

[0067] Specifically, the influence of input variables on output performance response is determined based on sensitivity indicators, and the parameter with the greatest influence on the sensitivity indicators corresponding to the radar system performance response is obtained to complete parameter dimensionality reduction.

[0068] To obtain high-precision parameter estimates, traditional parameter estimation algorithms are widely used in radar target parameter estimation, but most are based on the assumption that the radar system is precisely known. On the one hand, different radar systems have varying sensitivities to different parameters and exhibit different performance; therefore, the sensitivity of different parameters must be considered. Directly performing global sensitivity analysis to rank the importance of parameters would lead to some unimportant parameters being included in the calculation, resulting in redundancy and low efficiency. Therefore, a local sensitivity analysis method is first used to filter out parameters affecting radar system performance, ignoring parameters with low influence, thus improving the computational efficiency of sensitivity analysis. The local sensitivity analysis method involves making small changes to only one parameter near a fixed point in the parameter space, while keeping other parameters constant. The importance of a parameter is measured by the response change caused by a single parameter. By obtaining the influencing parameters through local sensitivity analysis, a global sensitivity analysis is then performed to rank the parameters by their degree of influence.

[0069] Specifically, the importance ranking of the influencing parameters is obtained according to the following steps:

[0070] The Sobol's sensitivity analysis method can be used to conduct a global sensitivity analysis, and the corresponding influencing parameters can be arranged in descending order of sensitivity indices.

[0071] Sobol's global sensitivity analysis aims to study the distribution of uncertainty in the model output response to uncertainty in the input, that is, to explore the source of uncertainty in the model output response, and to provide guidance for selecting a more reasonable and effective scheme to reduce the uncertainty in the model output response.

[0072] By prioritizing the importance of the parameters affecting each performance level, that is, by using the priority of each parameter relative to the performance of the phased array system, we can use this as the basis for performance analysis of the phased array radar system, better meet the design requirements, and achieve the goal of improving the model calculation efficiency of the phased array radar system.

[0073] Taking the maximum detection range and self-defense jamming range as examples, as shown in Table 3;

[0074] Table 3. Results of Global Sensitivity Analysis of Phased Array Radar System

[0075]

[0076] As shown in Table 3, the most important input parameter for the maximum detection range and the self-defense jamming range is obtained by arranging the multiple influencing parameters in descending order of sensitivity index. The most important input parameter for the maximum detection range is the receiving gain, and the most important input parameter for the self-defense jamming range is the jamming transmission gain. The most important input parameter for the entire phased array radar system is the receiving gain.

[0077] Compared to existing phased array radar model performance analysis, which suffers from numerous parameters and low computational efficiency, this invention first reduces the order of magnitude of parameter data intervals, with each set of input parameter samples containing multiple input parameters, thus transforming large intervals into smaller intervals and improving computational accuracy. Furthermore, by performing local sensitivity analysis on the phased array radar system, it obtains the influencing parameters and their dimensions corresponding to different performance characteristics, classifying and organizing the radar system performance parameters. This effectively reduces the dimensionality of the phased array radar system performance parameters while improving computational efficiency.

[0078] This invention explores the transmission of multi-source uncertainties in complex radar systems, and attempts to reduce or eliminate the impact of uncertainties on radar detection accuracy, thereby improving the reliability of phased array radar system simulation models.

[0079] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention is not limited to the specific combination of the above-described technical features, but also includes other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in this invention.

Claims

1. A method for performance calculation dimension reduction for a phased array radar system simulation model, characterized in that, The method comprises the following steps: Taking input parameter samples of a phased array radar system; each set of the input parameter samples comprises multiple parameter dimensions; Respectively, sequentially substituting each set of the input parameter samples into a phased array radar system simulation model to obtain corresponding radar system performance responses; Each time, only changing sample data of one of the parameter dimensions, and setting the remaining parameter dimensions as interval medians, and obtaining corresponding radar system performance responses through the phased array radar system simulation model; When all the parameter dimensions complete one calculation, all the influencing parameters of each radar system performance are obtained; when radar system performance responses obtained by multiple sample data of the same parameter dimension are inconsistent, the parameter dimension is the influencing parameter; the sensitivity of the parameter dimension is measured by the response change caused by a single parameter dimension, and whether the parameter dimension has an impact on the phased array radar system is determined; if the parameter dimension has an impact on the phased array radar system, the parameter dimension is an influencing parameter; Respectively, substituting sample data of the influencing parameters into the phased array radar system simulation model to obtain corresponding phased array radar system performance responses; Normalizing a plurality of sets of the input parameter samples and corresponding phased array radar system performance responses; Taking the normalized influencing parameters as input variables, and taking a sequence of phased array radar system performance responses corresponding to the normalized influencing parameters as output responses, a polynomial response surface surrogate model is established; If the polynomial response surface surrogate model meets a preset index, a ratio of a variance of each input parameter sample to a total variance is calculated to obtain a sensitivity index of different input parameter samples; According to the sensitivity index, an order of importance of parameters affecting the phased array radar system performance response is obtained.

2. The method of claim 1, wherein, Taking a plurality of sets of input parameter samples of a phased array radar system comprises the following steps: S1: obtaining a plurality of initial input parameters of the phased array radar system; each initial input parameter has a corresponding data interval.

3. The method of claim 2, wherein, After the polynomial response surface surrogate model is established and before the ratio of the variance of each input parameter sample to the total variance is calculated, the following steps are further included: If the polynomial response surface surrogate model does not meet the preset index, the above step S1 is performed.

4. The method of claim 1, wherein, Obtaining an order of importance of parameters affecting the phased array radar system performance comprises the following steps: Arranging the order of importance of parameters affecting the phased array radar system performance in descending order of the sensitivity index.

5. The method of claim 1, wherein, The radar system performance includes a maximum detection distance, a self-defense jamming distance, a detection probability, a distance accuracy, an azimuth angle and a pitch angle accuracy.

6. The method of claim 2, wherein, The initial input parameters include a signal / jamming frequency, a signal bandwidth, a pulse width, a pulse accumulation number, a radar transmission power, a transmission gain, a reception gain, a comprehensive loss, a suppression coefficient, a azimuth beam width, a pitch beam width, an angle error slope, a jamming bandwidth, a jamming power, and a jamming transmission gain.

7. The method of claim 1, wherein, The plurality of sets of input parameter samples of the phased array radar system are taken according to the following steps: A Latin hypercube sampling method is used to extract the plurality of sets of input parameter samples of the phased array radar system.

Images