CN-122022355-A - Task scheme generation system based on situation assessment

CN122022355ACN 122022355 ACN122022355 ACN 122022355ACN-122022355-A

Abstract

The application relates to the field of proposal generation, and particularly discloses a task proposal generation system based on situation assessment, which converts high-level task requirements and key characteristics of the current situation into formal specification problems, and the rule base and the case base are subjected to deep fusion and collaborative reasoning by combining an integrated reasoning engine, and the system can generate a scientific and feasible task scheme on the basis of fully understanding the complex battlefield environment. Compared with the traditional method, the scheme not only improves the intelligent level of task scheme generation, but also enhances the accuracy and adaptability of case matching through semantic embedded coding and other mechanisms, improves the diversity and tactical value of the scheme, and meets the urgent requirements of commanders for high-quality and multi-choice decision support in complex and changeable battlefield environments, thereby obviously improving the scientificity, timeliness and reliability of task command.

Inventors

ZHANG YANGMING
ZHANG LONG
YANG XIANGLI
SUN BIN
CHEN QIULI
ZHAO LINBANG
Jia Zichun
WANG JUN

Assignees

军事科学院系统工程研究院系统总体研究所

Dates

Publication Date: 20260512
Application Date: 20260206

Claims (8)

1. A task scheme generating system based on situation assessment is characterized by comprising a task requirement acquisition module, a current situation key feature acquisition module, an information conversion module, a preliminary task scheme generating module and a detail filling module, wherein the task requirement acquisition module is used for acquiring high-level task requirements input by a user, the current situation key feature acquisition module is used for acquiring current situation key features acquired by a situation awareness module, the information conversion module is used for converting the high-level task requirements and the current situation key features into formal specification problems, the preliminary task scheme generating module is used for inputting the formal specification problems, a task rule base and a task case base into an integrated reasoning engine to search solutions meeting the formal specification problems so as to obtain a preliminary task scheme skeleton, and the detail filling module is used for carrying out detail filling on the preliminary task scheme skeleton so as to obtain a detailed candidate task scheme.
2. The situation assessment based task solution generation system of claim 1, wherein the high-level task requirements include a target area, a time limit, an available type and amount of soldier, a forbidden area, and a performance index.
3. The situation assessment based mission solution generation system of claim 2, wherein the current situation critical features include enemy critical unit locations, terrain obstructions, weather conditions, and friend current status.
4. The situation-assessment-based task solution generation system according to claim 1, wherein the information conversion module includes a description information extraction unit configured to extract an initial state description, a target state description, and a constraint description from the high-level task requirements and the current situation key features, and a formal specification problem construction unit configured to construct the formal specification problem based on the initial state description, the target state description, and the constraint description.
5. The situation assessment-based task solution generating system according to claim 4, wherein the preliminary task solution generating module comprises a similar case segment retrieving unit configured to retrieve similar case segments from the task case library based on an initial state description in the formal specification problem by using the integrated inference engine, a preliminary solution skeleton generating unit configured to fuse the similar case segments into a current solution construction path to obtain a preliminary solution skeleton primitive after performing adaptive processing, and a checking unit configured to perform feasibility checking on the preliminary solution skeleton primitive based on the task rule library by using the integrated inference engine, and determine the preliminary solution skeleton primitive as the preliminary task solution skeleton after the feasibility checking is passed.
6. The situation evaluation-based task solution generating system according to claim 5, wherein the similar case segment retrieving unit comprises an initial state semantic coding subunit, a case segment semantic coding subunit, a semantic similarity calculating subunit and a similar case segment selecting subunit, wherein the initial state semantic coding subunit is used for performing semantic embedding coding on the initial state description to obtain initial state semantic embedded coding vectors, the case segment semantic coding subunit is used for performing semantic embedding coding on each case segment in the task case library to obtain a set of case segment semantic embedded coding vectors, the semantic similarity calculating subunit is used for calculating semantic similarity between the initial state semantic embedded coding vectors and each case segment semantic embedded coding vector in the set of case segment semantic embedded coding vectors, and the similar case segment selecting subunit is used for taking a case segment corresponding to the largest semantic similarity as the similar case segment.
7. The situation-assessment-based task solution generating system according to claim 6, wherein the semantic similarity calculating subunit is configured to obtain a conversion matrix based on the initial state semantic embedding encoding vector and the case segment semantic embedding encoding vector, limit the conversion matrix through local nonlinear conversion to generate a continuous diffusion analysis matrix, perform field dynamic mapping on the conversion matrix by the continuous diffusion analysis matrix to obtain a dynamic mapping matrix, perform semantic field adjustment on the case segment semantic embedding encoding vector based on the dynamic mapping matrix to obtain a case segment semantic embedding enhancement encoding vector, and calculate semantic similarity between the initial state semantic embedding encoding vector and the case segment semantic embedding enhancement encoding vector.
8. The situation assessment based task solution generating system according to claim 1, wherein the detail filling module is configured to identify variant points in the preliminary task solution skeleton, and fill the detail of each variant point to obtain the detailed candidate task solution, where details include specific unit allocation, precise time nodes, collaborative actions, and fire allocation.

Description

Task scheme generation system based on situation assessment Technical Field The application relates to the field of scheme generation, and more particularly relates to a task scheme generation system based on situation assessment. Background The situation assessment is rapidly and accurately carried out, and a scientific and feasible task scheme is generated, so that the method is a key for mastering initiative. The traditional scheme relies on commander experience to manually formulate, and the process is loaded down with trivial details, inefficiency, and easily receives subjective factor influence, is difficult to satisfy the demand of high strength, fast rhythm. In the face of an environment with high information density, how to effectively extract key situation information and support intelligent generation of a task scheme becomes a core problem to be solved. Therefore, an intelligent system which is based on real-time situation assessment and can automatically generate diversified and high-quality task schemes is constructed, and the intelligent system has important significance for improving the scientificity, timeliness and decision reliability of task command. To meet the urgent demands of modern tasks on intelligent decision systems, related studies have been conducted in academia and industry. The early system mostly adopts a rule-based reasoning method, and generates a scheme through a preset task regulations and expert knowledge, but is limited by the flexibility and completeness of a rule base, so that complex and changeable environments are difficult to deal with. Then, a case-based reasoning technology is introduced, a new scheme is generated by means of historical cases, however, the effect depends on a high-quality case base and a precise searching and adjusting mechanism, matching is difficult when facing a unique situation, the adjusting process is complex, and the innovation of the scheme is insufficient. In order to further improve the level of intellectualization, part of research attempts to fuse rules and case reasoning to construct a hybrid reasoning system, but the deep integration difficulty still faces in the aspect of knowledge representation and reasoning coordination, and the real advantage complementation is difficult to realize. In addition, existing systems represent a limited number of solution diversity, multiple outputs are single "optimal solutions", and directors often need multiple candidate solutions with distinct and distinct task features to cope with uncertainty, which is difficult to meet by means of parameter adjustment or randomization methods alone. Therefore, how to construct an intelligent system that can deeply fuse multi-source knowledge, fully understand situations, and generate diversified task schemes remains an important challenge in the current field. Disclosure of Invention The present application has been made to solve the above-mentioned technical problems. According to one aspect of the application, a task scheme generating system based on situation assessment is provided, which comprises a task requirement acquisition module, a current situation key feature acquisition module, an information conversion module, a preliminary task scheme generating module and a detail filling module, wherein the task requirement acquisition module is used for acquiring high-level task requirements input by a user, the current situation key feature acquisition module is used for acquiring current situation key features acquired by a situation awareness module, the information conversion module is used for converting the high-level task requirements and the current situation key features into formal specification problems, the preliminary task scheme generating module is used for inputting the formal specification problems, a task rule library and a task case library into an integrated reasoning engine to search solutions meeting the formal specification problems so as to obtain a preliminary task scheme skeleton, and the detail filling module is used for carrying out detail filling on the preliminary task scheme skeleton so as to obtain detailed candidate task schemes. Compared with the prior art, the task scheme generating system based on situation assessment, provided by the application, has the advantages that the rule base and the case base are fused through the integrated reasoning engine, the limitation of a single reasoning method is overcome, the semantic embedded coding and dynamic state mapping technology improves the accuracy of case retrieval and the adaptability to unique situations, the problem of difficult traditional case reasoning matching is solved, the feasibility checking mechanism of the rule base compensates the complexity of case adjustment, and the scheme basic rationality is ensured. Meanwhile, a detail filling mechanism driven by a variation point is innovatively introduced into the system, and differentiated details such as specific unit