English translation activity course is of great significance to cultivate students’ English translation level. In the context of multisensor data fusion, how to effectively carry out English translation activity course in colleges and universities has become an important topic. The educational value and intellectual property of advanced Intelligent English translation activity course are analyzed. From multisensor data fusion and the improvement of translation, translators psychological changes of the boot and prominent features, English translation activity and translation, the generality of the multisensor data fusion of multisensor data fusion personalization features this four aspects, which is under the background of the sensor data fusion of the practice for college English translation activity. Firstly, the theory of data fusion estimation is elaborated, and various data fusion structures in multisensor systems are summarized. Then, the data fusion estimation model based on Kalman filter is established, and the Kalman filtering algorithms of centralized, sequential, parallel, and joint structures are given, respectively. Simulation experiments are carried out on the algorithms. Experimental results show that the estimation accuracy of the system can be improved by multisensor data fusion. Then a rule-based lexical analyzer is designed. Combined with the system model, a rule-based lexical analyzer and a comprehensive dictionary, the linguistic knowledge source throughout the whole machine translation process, are researched and designed. A hashing algorithm for lexical retrieval is designed, and various rules and data structures related to the lexical analyzer are described in formal language. The analysis algorithms of morphological preprocessing, morphological analysis, unincluded word processing, phrase analysis, and part of speech tagging are introduced.