The stable operation of key electrical equipment is more and more important for guaranteeing the safe and reliable production of offshore oil and gas platforms, and with the development of online monitoring, pattern recognition, computer information processing technology, etc., it has become an inevitable trend for the electrical equipment to change from the current planned maintenance to condition maintenance. In order to realize the effective diagnosis of electrical equipment faults in offshore oil and gas fields, this paper proposes an offshore oil and gas field equipment fault diagnosis model based on deep reinforcement learning. Multi-source heterogeneous data of offshore oil and gas field electrical equipment are collected, and the fault features are selected and extracted by adaptive NLM algorithm, and combined with the cubic spline interpolation algorithm to fill in the data in order to ensure the completeness of the fault data. Then, deep reinforcement learning is combined with deep domain adaptive networks to establish a cross-condition fault diagnosis model for electrical equipment in offshore oil and gas fields. The fault diagnosis accuracy of electrical equipment in offshore oil and gas fields designed in this paper can reach up to 98.95%, and the diagnosis accuracy of the model in this paper is improved by 10~32 percentage points compared with shallow migration learning. Therefore, the application of deep reinforcement learning technology to the fault diagnosis of electrical equipment in offshore oil and gas fields can detect the fault conditions of electrical equipment in time and maintain the stable operation of electrical equipment in offshore oil and gas fields.