Heterogeneous multiprocessor computing systems are widely used in many fields because of their powerful computational abilities and productive parallel processing performance. Nevertheless, their excellent performance is accompanied by the problem of rapidly growing energy consumption. In this work, we concentrate on the issue of minimizing workflow scheduling length under energy consumption limitations in heterogeneous multiprocessor computing systems. First, we construct a model of processor workflow and energy consumption, and regard the scheduling problem under the limitations of energy consumption as an optimization problem whose core objective is to maximize the compression of the scheduling length of the workflow. Then, we present a new scheduling algorithm depending on task characteristics, which considers the two key characteristics of the task’s execution time and energy consumption on the processor when allocating energy, and allocates energy to the task in a more rational way. Concurrently, we apply an earliest-completion-time strategy to assign tasks to optimal execution frequencies and processors. Finally, we verify the reliability and stability of the proposed algorithm through real-world workflow experiments with randomly generated workflows.