Integrated Intelligence Evaluation for Unmanned Systems Using Dempster-Shafer Fusion

To address the limitations of existing evaluation methods in quantifying fuzziness and randomness for unmanned systems’ intelligence level, this study proposes an assessment framework integrating Dempster-Shafer (D-S) evidence fusion and cloud models. A hierarchical indicator system is constructed, covering autonomy, collaboration, and learning capabilities. Subjective weights are calculated via the Analytic Hierarchy Process (AHP), while objective weights are derived from the entropy method. The D-S evidence theory fuses these weights, resolving conflicts and uncertainty between subjective and objective perspectives. Cloud models quantify the fuzziness and randomness of evaluation. Case validation demonstrates that the framework effectively identifies system vulnerabilities and intelligence deficiencies, providing a dynamic and quantifiable methodology for assessing intelligent unmanned systems.