The rise of intelligent vehicular applications has amplified the demand for high-performance computing, addressing the constraints of limited onboard computational resources through Vehicular Fog Computing (VFC). VFC offers low latency and swift response times but also presents security challenges. Blockchain technology can secure transactions in VFC, but it consumes significant resources. In addition, inefficient task scheduling strategies can increase task latency, reduce service quality, and raise the number of blockchain transactions, worsening resource consumption and straining the Internet of Vehicles (IoV) systems. To this end, we propose IADE-DPSAC, a task scheduling strategy based on heuristic algorithms and deep reinforcement learning, which jointly optimizes fog computing task offloading and resource allocation, significantly improving computational efficiency and reducing blockchain resource consumption.