A new λ-fuzzy quasi product operator and its fuzzy probability integral are proposed and applied to the evaluation of performance state of pulverized coal rotary feeder. Based on the collection of relevant data in the mechanical laboratory， the AHP-improved entropy weight method is used to determine the weight of each index in the evaluation process， which overcomes the shortcomings of the traditional entropy weight method. The fuzzy probability value calculated by the above method combined with the maximum probability principle is used to evaluate the performance state of the pulverized coal rotary feeder. Finally， the results are compared with those obtained by the fuzzy comprehensive evaluation method. The comparison results show that the evaluation results of the λ-fuzzy quasi probability integral method are more accurate， and the feeders can be sorted， which overcomes the shortcomings of the traditional fuzzy comprehensive evaluation results that can only judge the excellent and poor performance of the device.