The article describes the structure of a fuzzy knowledge base, which is used in an information system for analyzing the state of complex technical systems. The knowledge base is a fuzzy Web Ontology Language (OWL) ontology. In addition, the knowledge base includes a set of Semantic Web Rule Language (SWRL) rules compiled by an expert and integrated with the ontology. Triangular and trapezoidal membership functions are used. The parameters of the membership functions are specified in the datatype properties of the ontology. The algorithm for integrating the fuzzy OWL ontology and the set of SWRL rules is used to detect abnormal values ​​using the example of analyzing the state of a data storage system. The article provides a brief description of the software system and the experiments conducted. The scientific novelty of the study is the development of a new approach to monitoring the state of complex technical systems, which is based on the integration of the Long short-term memory (LSTM) and a knowledge base consisting of a fuzzy ontology and a set of SWRL rules. Experiments were conducted to detect anomalies of sent datagrams in a data storage server. Anomalies were detected using 4 approaches: recurrent neural networks (RNNs), convolutional neural network (CNN), LSTM and a method using a fuzzy knowledge base. The proposed approach showed a slightly better value than LSTM with an accuracy value of 0.86.