In order to achieve low-cost production of customized components, production scheduling needs to analyze the reliability and stability of equipment on the basis of processing efficiency and quality for integrating appropriate equipment maintenance strategy. For machining, the actual process also needs to consider the deteriorating effect brought by cutters’ wear. In order to solve the single machine scheduling problem with state-strategy maintenance and step-deteriorating jobs, this paper analyzes continuous fault detection and discrete fault detection, and formulates the single machine scheduling model with minimization of tardiness and maintenance costs when considering piecewise linear deterioration and maintenance unavailable time windows. Through the normalization of decision-making conditions, the advantages and disadvantages of maintenance strategies are discussed. Based on single-factor and two-factor tuning parameter analysis, the impact of related parameter changes on scheduling optimization is analyzed. The computational results show that discrete fault detection easily causes over-maintenance or under-maintenance. Once adopting normalization in the two strategies better results can be obtained, but cannot offset the penalty time of deterioration. In that situation, the target cost is not always the smallest.