A method for correction and diagnosis of measurement errors is proposed where unlike in the "crisp" case a faulty degree of each measuring instrument between 0 and 1 is introduced. This allows the correction procedure to be performed smoothly for an arbitrary number of wrong instruments within the range of faulty degrees between 0 and 1. The proposed method for measurement corrections called further Soft Computing Method has lead also to a simplified procedure for diagnosis of faulty instruments. A test and an industrial example have shown the ability of the proposed method for implementation into industrial control systems for real time measurement correction and diagnosis.