Comprehensive radio frequency system overall performance evaluation method based on BP neural network

Abstract

The invention discloses an integrated radio frequency system overall performance evaluation method based on a BP neural network. The method comprises the following steps: 1) establishing an integrated radio frequency system overall performance evaluation index system; 2) collecting performance indexes of the technical performance layer, and normalizing and quantifying all the indexes of the technical performance layer; 3) establishing a BP neural network model for performance evaluation of the integrated radio frequency system; and 4) according to the output of the efficiency evaluation model, determining the efficiency level of the overall efficiency evaluation of the integrated radio frequency system as the output result of the overall efficiency evaluation.According to the invention, a new overall efficiency evaluation index system of the integrated radio frequency system is provided, the adaptive efficiency evaluation model is constructed, and quantitative analysis of the overall efficiency of the integrated radio frequency system can be carried out.

Description

Technical Field

[0001] The present invention relates to a radio frequency equipment performance evaluation technology, and in particular to a comprehensive radio frequency system overall effectiveness evaluation method based on a BP neural network. Background Art

[0002] An integrated RF system refers to the integration of RF equipment with different functions such as detection, reconnaissance, electronic countermeasures, and communication under typical situations. It has the ability to concentrate resources and intelligently control and dispatch them. In various typical scenarios and application modes, it optimizes resource utilization and combat processes based on the usage processes of different functions and the mutual resource constraints between the functions of the integrated system, and improves the overall effectiveness of the system. An integrated RF system generally adopts a modular, open, and reconfigurable RF sensor system architecture, combined with functional control and resource dispatching algorithms and software, to achieve system integration of multiple functions such as radar, electronic warfare and communication, navigation, and identification.

[0003] The overall effectiveness of an integrated RF system refers to the degree to which the system meets a set of expected mission requirements under specified conditions and within a specified time. The overall effectiveness of an integrated RF system is not only a multi-dimensional description of the system's comprehensive capabilities, but also the basic basis for the demonstration, development, planning and configuration of an integrated RF system. The overall effectiveness evaluation of an integrated RF system is of great significance to the development, improvement and comprehensive use of integrated systems for ships, air-based platforms and other platforms. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method for evaluating the overall effectiveness of an integrated radio frequency system based on a BP neural network in view of the defects in the existing integrated radio frequency system effectiveness evaluation technology.

[0005] The technical solution adopted by the present invention to solve the technical problem is: a method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network, comprising the following steps: 1) Establish an overall performance evaluation index system for integrated RF systems; The comprehensive RF system performance evaluation index system includes: The overall effectiveness layer of the integrated radio frequency system is the overall effectiveness of the integrated radio frequency system in carrying out various integrated tasks such as detection, electronic countermeasures, reconnaissance, and communication; The overall effectiveness layer of the integrated radio frequency system includes the following functional mission effectiveness: detection effectiveness, electronic countermeasure effectiveness, reconnaissance effectiveness and communication effectiveness; in, Detection effectiveness refers to the technical capabilities of detection effectiveness, including: detection signal transmission capability and signal processing capability; Electronic countermeasure effectiveness refers to the technical capabilities of electronic countermeasure effectiveness, including: electronic countermeasure signal emission capability and signal guidance capability; Reconnaissance effectiveness refers to the technical capabilities of reconnaissance effectiveness, including: signal reception capability and signal processing capability; Communication performance refers to the technical capabilities of communication performance, including: network capacity, latency parameters, and transmission quality; Technical performance layer, including the following performance indicators: Performance indicators of detection signal transmission capability: maximum operating distance D1, detection accuracy D2; Performance indicators of signal processing capability: signal processing data rate D3, resolution D4, multi-target capacity D5, and anti-interference capability D6; Performance indicators of electronic countermeasure signal transmission capability: maximum transmission power D7, instantaneous interference bandwidth D8, minimum interference distance D9; Performance indicators of signal guidance capability: interference guidance time D10, azimuth aiming accuracy D11, multi-target interference capability D12; Performance indicators of signal detection capability: detection sensitivity D13, interception probability D14; Performance indicators of signal processing capability: frequency measurement capability D15, direction finding capability D16, simultaneous multi-target tracking capacity D17, signal sorting capability D18; Performance indicators of network capacity: number of network nodes D19; Performance indicators of delay parameters: synchronization time D20, transmission delay D21; Performance indicators of transmission quality: transmission rate D22, bit error rate D23, frame loss rate D24; 2) Collect performance indicators of the technical performance layer, model and normalize and quantify all indicators of the technical performance layer, namely the technical performance evaluation indicator layer; The performance indicators of the technical performance layer are directly obtained through system equipment attribute parameters, obtained through system equipment attribute parameter representation, or obtained through modeling; Each evaluation index takes a value between [0,1]. When the evaluation index value is positively correlated with the comprehensive performance, the sample Normalized value for: ; When the evaluation index value is negatively correlated with the comprehensive performance, the sample Normalized value for: .

[0006] 3) Establish a BP neural network model for comprehensive RF system performance evaluation; The number of nodes in the input layer and the hidden layer is determined according to the performance evaluation index, and the training process of the BP neural network model is started according to the training samples. After the training results converge, the performance evaluation model based on the BP neural network is obtained.

[0007] 4) According to the output of the model, determine the performance level of the comprehensive performance evaluation of the integrated RF system as the output result of the comprehensive performance evaluation.

[0008] According to the above scheme, in the step 2), the performance indicators D1 to D5, D7 to D17, and D19 to D24 are obtained through the system equipment attribute parameters; among them, the multi-target interference capability D12 can be characterized by the number of simultaneous multi-target interference batches, the frequency measurement capability D15 is characterized by the frequency measurement accuracy error parameter of the reconnaissance function, the direction finding capability D16 is characterized by the direction finding error parameter of the reconnaissance function, and the multi-target tracking capacity D17 is characterized by the number of batches of multi-targets tracked simultaneously by the reconnaissance function; other performance indicators are directly obtained through the system equipment attribute parameters.

[0009] According to the above scheme, in step 2), the performance indicator anti-interference capability D6 obtains the evaluation level of the anti-interference capability based on the neural network model or expert system with the frequency conversion bandwidth, the number of types of anti-interference methods, and the anti-interference capability as input.

[0010] According to the above scheme, in step 2), the signal sorting capability D18 is measured by the ability to identify and eliminate false targets and to perform multi-target fusion processing, and the capability level of the signal sorting capability is obtained through an expert system or a neural network model.

[0011] According to the above scheme, in step 2), all indicators of the technical performance layer are normalized and quantified; Each evaluation index takes a value between [0,1]. When the evaluation index value is positively correlated with the comprehensive performance, the sample Normalized value for: ; When the evaluation index value is negatively correlated with the comprehensive performance, the sample Normalized value for: .

[0012] According to the above scheme, in step 3), the BP neural network model is as follows: The number of input layer nodes n is the number of indicators involved in performance evaluation; The number of hidden layer nodes h is determined according to the following formula:

[0013] Among them, h is the number of hidden layer nodes, n is the number of input layer nodes, and m is the number of output layer nodes. a is a random constant between 1 and 10; The training function expression of the BP neural network for comprehensive RF system performance evaluation is as follows:

[0014] in, is the neural network weight parameter, is the learning rate; The loss function L used in training adopts the cross entropy error loss function.

[0015] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the overall effectiveness evaluation method of an integrated radio frequency system based on a neural network in the above-mentioned scheme are implemented.

[0016] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the overall effectiveness evaluation method of an integrated radio frequency system based on a neural network in the above-mentioned scheme is implemented.

[0017] The beneficial effects produced by the present invention are: 1. The present invention proposes a new comprehensive radio frequency system overall performance evaluation index system. According to the task requirements and specific usage processes, an adaptive performance evaluation model is constructed to perform quantitative analysis of the overall performance of the comprehensive radio frequency system. 2. Based on the model of the present invention, when radio frequency resources are scarce or there are conflicts in the use of various functions, in order to achieve the best overall performance, accurate allocation of radio frequency resources can be carried out according to the comprehensive performance evaluation results of the integrated radio frequency system to meet the coordinated use of various detection, electronic countermeasures, reconnaissance, communication and other functions, ensuring that various functions work simultaneously and with optimal compatibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which: Figure 1 is a diagram of an overall performance evaluation index system of a comprehensive radio frequency system according to an embodiment of the present invention; Figure 2 Schematic diagram of the BP neural network structure for comprehensive radio frequency system performance evaluation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0020] like Figure 1As shown, a method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network comprises the following steps: 1) Establish an overall performance evaluation index system for integrated RF systems; The comprehensive RF system performance evaluation index system includes: The overall effectiveness layer of the integrated radio frequency system is the overall effectiveness of the integrated radio frequency system in carrying out various integrated tasks such as detection, electronic countermeasures, reconnaissance, and communication; The overall effectiveness layer of the integrated radio frequency system includes the following functional mission effectiveness: detection effectiveness, electronic countermeasure effectiveness, reconnaissance effectiveness and communication effectiveness; in, Detection effectiveness refers to the technical capabilities of detection effectiveness, including: detection signal transmission capability and signal processing capability; Electronic countermeasure effectiveness refers to the technical capabilities of electronic countermeasure effectiveness, including: electronic countermeasure signal emission capability and signal guidance capability; Reconnaissance effectiveness refers to the technical capabilities of reconnaissance effectiveness, including: signal reception capability and signal processing capability; Communication performance refers to the technical capabilities of communication performance, including: network capacity, latency parameters, and transmission quality; Technical performance layer, including the following performance indicators: Performance indicators of detection signal transmission capability: maximum operating distance D1, detection accuracy D2; Performance indicators of signal processing capability: signal processing data rate D3, resolution D4, multi-target capacity D5, and anti-interference capability D6; Performance indicators of electronic countermeasure signal transmission capability: maximum transmission power D7, instantaneous interference bandwidth D8, minimum interference distance D9; Performance indicators of signal guidance capability: interference guidance time D10, azimuth aiming accuracy D11, multi-target interference capability D12; Performance indicators of signal detection capability: detection sensitivity D13, interception probability D14; Performance indicators of signal processing capability: frequency measurement capability D15, direction finding capability D16, simultaneous multi-target tracking capacity D17, signal sorting capability D18; Performance indicators of network capacity: number of network nodes D19; Performance indicators of delay parameters: synchronization time D20, transmission delay D21; Performance indicators of transmission quality: transmission rate D22, bit error rate D23, frame loss rate D24; 2) Collect the performance indicators of the technical performance layer and normalize and quantify all the indicators of the technical performance layer, that is, the technical performance evaluation indicator layer; The performance indicators of the technical performance layer are directly obtained through system equipment attribute parameters, obtained through system equipment attribute parameter representation, or obtained through modeling; Most performance indicators can be obtained through system attribute parameters, such as the performance indicators D1, D2, D3, D4, and D5 of the detection function, the performance indicators D7, D8, D9, D10, D11, and D12 of the electronic countermeasure function, the performance indicators D13, D14, D15, D16, and D17 of the reconnaissance function, and the performance indicators D19, D20, D21, D22, D23, and D24 of the communication function. Among them, the multi-target interference capability D12 can be characterized by the parameter of the number of simultaneous multi-target interference batches, the frequency measurement capability D15 can be characterized by the frequency measurement accuracy error parameter of the reconnaissance function, the direction finding capability D16 can be characterized by the direction finding error parameter of the reconnaissance function, and the simultaneous tracking of multiple targets capacity D17 can be characterized by the number of simultaneous tracking of multiple targets batches parameter of the reconnaissance function.

[0021] A small number of performance indicators can be obtained based on the system qualitative parameters combined with expert evaluation or modeling. For example, the performance indicator of the detection function, the anti-interference capability D6, can be evaluated through a neural network model or an expert system scoring method based on system characteristics such as frequency conversion bandwidth, the number of anti-interference methods, and anti-clutter capability (for example, if the anti-interference capability index of a system with 10 anti-interference methods is 1, then the capability index of a system with only 8 anti-interference methods is 0.8, or an expert evaluation can be performed based on the anti-clutter capability parameters), and the signal sorting capability D18 of the reconnaissance function can be measured by the ability to identify and eliminate false targets and to perform multi-target fusion processing, and can be given through an expert system or a neural network model.

[0022] Then the obtained indicators are normalized.

[0023] Each evaluation index takes a value between [0,1]. When the evaluation index value is positively correlated with the comprehensive performance, the sample Normalized value for: ; When the evaluation index value is negatively correlated with the comprehensive performance, the sample Normalized value for: .

[0024] 3) Establish a BP neural network model for comprehensive RF system performance evaluation; The number of nodes in the input layer and the hidden layer is determined according to the performance evaluation index, and the training process of the BP neural network model is started according to the training samples. After the training results converge, the performance evaluation model based on the BP neural network is obtained.

[0025] The number of input layer nodes n is the number of indicators involved in performance evaluation; The number of hidden layer nodes h is determined according to the following formula:

[0026] Among them, h is the number of hidden layer nodes, n is the number of input layer nodes, and m is the number of output layer nodes. a is a random constant between 1 and 10; The training function expression of the BP neural network for comprehensive RF system performance evaluation is as follows:

[0027] in, is the neural network weight parameter, is the learning rate; The loss function L used in training adopts the cross entropy error loss function; After repeated learning and training, the neuron parameters are continuously adjusted and the error is continuously reduced until the accuracy requirement is met or the pre-set maximum number of learning times is reached, then the training is stopped.

[0028] The BP neural network structure for comprehensive performance evaluation of integrated radio frequency system is as follows: Figure 2 As shown in the figure, a training sample consists of the normalized values ​​of 24 performance evaluation index values ​​of an alternative plan and the corresponding quantized value of the comprehensive performance evaluation level. The training sample set has a total of 200 sets of data, which are obtained from 200 alternative use plans. The number of input layer nodes of the model is 24, the number of output layer nodes is 5, and the number of hidden layer nodes is 10; the activation function is ReLU; the training function uses the average stochastic gradient descent ASGD function, the error loss function is the cross entropy error loss function, the learning rate is 0.001, and the maximum number of training rounds is 2000.

[0029] 4) According to the output of the model, determine the performance level of the overall performance evaluation of the integrated RF system as the output result of the comprehensive performance evaluation.

[0030] In this example, the performance level results of the comprehensive performance of the integrated RF system are divided into five levels: "excellent", "good", "qualified", "poor" and "very poor". The quantitative value range of each performance level is shown in Table 1.

[0031] Table 1 Quantitative values ​​of performance levels

[0032] About the effect of the model: For example, the performance index value vector of the integrated RF systems D1 to D24 is (0.85, 0.06, 0.17, 0.28, 0.66, 0.18, 0.89, 0.04, 0.57, 0.55, 0.77, 0.41, 0.62, 0.15, 0.54, 0.33, 0.09, 0.76, 0.67, 0.68, 0.49, 0.01, 0.26, 0.80). Based on this, the experts assess that the overall performance of the integrated RF system is “poor”. At the same time, the group of performance index values ​​is input into the trained comprehensive RF system comprehensive performance evaluation model based on BP neural network, and arranged in the order of "very poor", "poor", "qualified", "good" and "excellent". The probability vector composed of the corresponding comprehensive performance level probability output by the model after calculation is (0.0043, 0.9717, 0.0240, 0.0000, 0.0000), and the comprehensive performance is "poor", which is consistent with the expert evaluation. The comprehensive performance evaluation results of the comprehensive RF system based on BP neural network and the direct judgment evaluation results of experts are well consistent.

[0033] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network, characterized in that: The following steps are involved: 1) Establish an overall performance evaluation index system for integrated RF systems; The comprehensive RF system performance evaluation index system includes: The overall effectiveness layer of the integrated radio frequency system is the overall effectiveness of the integrated radio frequency system in carrying out various integrated tasks such as detection, electronic countermeasures, reconnaissance, and communication; The overall effectiveness layer of the integrated radio frequency system includes the following functional mission effectiveness: detection effectiveness, electronic countermeasure effectiveness, reconnaissance effectiveness and communication effectiveness; in, Detection effectiveness refers to the technical capabilities of detection effectiveness, including: detection signal transmission capability and signal processing capability; Electronic countermeasure effectiveness refers to the technical capabilities of electronic countermeasure effectiveness, including: electronic countermeasure signal emission capability and signal guidance capability; Reconnaissance effectiveness refers to the technical capabilities of reconnaissance effectiveness, including: signal reception capability and signal processing capability; Communication performance refers to the technical capabilities of communication performance, including: network capacity, latency parameters, and transmission quality; Technical performance layer, including the following performance indicators: Performance indicators of detection signal transmission capability: maximum operating distance D1, detection accuracy D2; Performance indicators of signal processing capability: signal processing data rate D3, resolution D4, multi-target capacity D5, and anti-interference capability D6; Performance indicators of electronic countermeasure signal transmission capability: maximum transmission power D7, instantaneous interference bandwidth D8, minimum interference distance D9; Performance indicators of signal guidance capability: interference guidance time D10, azimuth aiming accuracy D11, multi-target interference capability D12; Performance indicators of signal detection capability: detection sensitivity D13, interception probability D14; Performance indicators of signal processing capability: frequency measurement capability D15, direction finding capability D16, simultaneous multi-target tracking capacity D17, signal sorting capability D18; Performance indicators of network capacity: number of network nodes D19; Performance indicators of delay parameters: synchronization time D20, transmission delay D21; Performance indicators of transmission quality: transmission rate D22, bit error rate D23, frame loss rate D24; 2) Collect the performance indicators of the technical performance layer and normalize and quantify all indicators of the technical performance layer, that is, the performance indicator modeling and evaluation layer; The indicators of the technical performance layer are directly obtained through system equipment attribute parameters, obtained through system equipment attribute parameter representation, or obtained through modeling; 3) Establish a BP neural network model for comprehensive RF system performance evaluation; The number of nodes in the input layer and the hidden layer is determined according to the performance evaluation index, and the training process of the BP neural network model is started according to the training samples. After the training results converge, the performance evaluation model based on the BP neural network is obtained; 4) According to the output of the performance evaluation model, determine the performance level of the comprehensive performance evaluation of the integrated radio frequency system as the output result of the comprehensive performance evaluation.

2. The method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network according to claim 1, characterized in that: In the step 2), the performance indicators D1 to D5, D7 to D17 and D19 to D24 are obtained through system equipment attribute parameters; among them, the multi-target interference capability D12 is characterized by the number of simultaneous multi-target interference batches, the frequency measurement capability D15 is characterized by the frequency measurement accuracy error parameter of the reconnaissance function, the direction finding capability D16 is characterized by the direction finding error parameter of the reconnaissance function, and the multi-target tracking capacity D17 is characterized by the number of simultaneous multi-target tracking batches parameter of the reconnaissance function; other performance indicators are directly obtained through system equipment attribute parameters.

3. The method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network according to claim 1, characterized in that: In the step 2), the performance indicator anti-interference capability D6 obtains the evaluation level of the anti-interference capability according to the neural network model or expert system with the frequency conversion bandwidth, the number of types of anti-interference methods, and the anti-clutter capability as input.

4. The method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network according to claim 1, characterized in that: In the step 2), the signal sorting capability D18 is measured by the ability to identify and eliminate false targets and to perform multi-target fusion processing, and the capability level of the signal sorting capability is obtained through an expert system or a neural network model.

5. The method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network according to claim 1, characterized in that: In the step 2), all indicators of the technical performance layer are normalized and quantified; Each evaluation index takes a value between [0,1]. When the evaluation index value is positively correlated with the comprehensive performance, the sample Normalized value for: ； When the evaluation index value is negatively correlated with the comprehensive performance, the sample Normalized value for: 。 6. The method for evaluating the overall effectiveness of a comprehensive radio frequency system based on a BP neural network according to claim 1, characterized in that: In step 3), the BP neural network model is as follows: The number of input layer nodes n is the number of indicators involved in performance evaluation; The number of hidden layer nodes h is determined according to the following formula: Among them, h is the number of hidden layer nodes, n is the number of input layer nodes, and m is the number of output layer nodes. a is a random constant between 1 and 10; The training function expression of the BP neural network for comprehensive RF system performance evaluation is as follows: in, is the neural network weight parameter, is the learning rate; The loss function L used in training adopts the cross entropy error loss function.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method for evaluating the overall effectiveness of an integrated radio frequency system based on a BP neural network according to any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the overall effectiveness evaluation method of an integrated radio frequency system based on a BP neural network as described in any one of claims 1 to 6 is implemented.

Images