Fitting the Hill Equation to Experimental Data

A procedure is presented for fitting the hyperbolic Hill relationship to force-velocity curves measured on muscle. The procedure combines linear least-squares estimation of two parameters with iterative estimation of the third parameter in the Hill model. The method was tested on curves generated with known parameters, with the addition of white noise, and on experimentally determined curves. It is found to give a better fit than the simple linear method when noise is present in the data.