The object of study is the problem of creating a relatively simple diagnostic model of a diesel engine for an autonomous power plant in terms of the number of required input parameters. The subject of the study is the relationship between the selected set of input parameters of the diagnostic model of a diesel engine and the adequacy of the response of the resulting model to real emergency situations. As a result of the study, a methodology for synthesising a diagnostic graph model of a diesel engine of an autonomous power plant has been developed. The resulting model allows to achieve control of diesel engine performance and adjustment of the relevant parameters to achieve the required operating condition or localisation of an emergency. To build a diagnostic system, a method of synthesising a model for diagnostics based on the use of directed graph theory approaches is proposed. This approach makes it possible, with relative technical simplicity of implementation, to obtain results that correspond to the real state of operation of the autonomous power supply system. A methodology for selecting the output parameters of the diagnostic model of a diesel engine is proposed, which determines the algorithm for forming the minimum required set of input parameters of the engine. As variants of this set of parameters, power, diesel engine crankshaft speed, fuel consumption, and a group of thermal engineering parameters of the power plant functioning are chosen. It is shown that such a set of parameters makes it possible to link the diesel diagnostic system with its performance indicators. The scope of application of the results of the study is the development of power engineering with diesel engines, where the proposed diagnostic approaches are expected to improve the quality and efficiency of ensuring uninterrupted operation.