An Evaluation Method of Combat Service Operation Performance for Surface-to-Air Missile Based on FNN

Gang Li, Dapeng Li, Yanyan Zhou, Linwei Xu, Jifeng He2012 · DOI

Abstract

Firstly, the implication and purpose of the Combat Service Performance Evaluation for Air Defense Missile was elaborated in this paper. Then the Fuzzy Neural Network (FNN) Evaluating Method was put forward to apply to the combat service performance evaluation. Secondly, an evaluation index system was built on the basis of the characteristics of the combat service operator and the combat service process. At last, the correctness and validity were approved though the simulation results.

Introduction

Evaluating the effectiveness of a combat crew in surface-to-air missile (SAM) systems is a complex task. These systems require seamless coordination between command vehicles, search radars, and launch platforms. Traditional evaluation methods often suffer from subjectivity, where instructors may apply inconsistent criteria, and complexity, as combat service involves a mix of quantitative data (e.g., target lock time) and qualitative traits (e.g., psychological stability). To address these issues, this paper proposes a comprehensive evaluation model using a Fuzzy Neural Network (FNN) to provide objective, accurate, and rapid performance grading.

Methodology

The Evaluation Index System

A robust evaluation requires a structured breakdown of combat performance. The researchers established a five-layer hierarchical index system to transform high-level objectives into granular, measurable metrics. This hierarchy spans from the overall target (combat performance) down to specific operational factors like "speed of target" and "psychology diathesis."

The hierarchical evaluation index system used to decompose combat performance into measurable metrics.

The system aggregates scores from four key roles: the commander, and the operators of the C-vehicle (command and control), S-vehicle (search), and F-vehicle (emission/launch). By structuring inputs this way, the model ensures that the neural network evaluates relevant military competencies rather than arbitrary correlations.

Fuzzy Neural Network (FNN) Architecture

The core of the evaluation engine is a Backpropagation (BP) Fuzzy Neural Network. This hybrid approach leverages fuzzy logic to handle qualitative human judgments—converting labels like "Excellent" or "Good" into mathematical membership degree vectors. These vectors are then processed by a multi-layer neural network consisting of an input layer (4 neurons representing the main criteria), a fuzzy quantization layer (12 neurons), a hidden layer (13 neurons), and an output layer (3 neurons). This architecture allows the system to model the non-linear relationships inherent in complex combat scenarios.

Results and Discussion

The model was validated using simulation data from 10 groups of combat service operations. Eight groups were used for training, while two were reserved for testing. The "ideal" outputs for training were generated using the Fuzzy Analytic Hierarchy Process (FAHP) to ensure the network learned from established expert logic.

The network demonstrated efficient learning, reaching convergence after 471 training steps. As shown in the results table, the error between the ideal expert scores and the FNN's actual outputs was remarkably low.

Comparison between ideal expert scores and the FNN's actual outputs across training and testing samples.

For instance, in test sample 10, the ideal score was 82.8506, and the FNN predicted 83.2388, representing a minimal error of 0.0047. This high precision, combined with the speed of automated processing, confirms the FNN's validity as a tool for real-time training feedback.

Conclusion

The proposed FNN-based evaluation method successfully bridges the gap between subjective expert judgment and objective mathematical modeling. By utilizing a comprehensive index system and a fuzzy neural architecture, the system provides accurate qualitative and quantitative performance assessments. While the current study is limited by a small sample size and simulated data, it lays the groundwork for more robust, automated military training evaluation systems.

Summary updated

Abstract

Firstly, the implication and purpose of the Combat Service Performance Evaluation for Air Defense Missile was elaborated in this paper. Then the Fuzzy Neural Network (FNN) Evaluating Method was put forward to apply to the combat service performance evaluation. Secondly, an evaluation index system was built on the basis of the characteristics of the combat service operator and the combat service process. At last, the correctness and validity were approved though the simulation results.
BibTeX
@misc{li2012evaluation,
  author = {Gang Li and Dapeng Li and Yanyan Zhou and Linwei Xu and Jifeng He},
  title = {{An Evaluation Method of Combat Service Operation Performance for Surface-to-Air Missile Based on FNN}},
  year = {2012},
  doi = {10.2991/iccia.2012.96},
}
Figure