Abstract: The adaptive crossover factor and mutation factor are used to enhance the global search ability and improve the convergence speed. The optimal value of the parameter M in the Stearns−Noechel model is determined by the method of loop iteration. The new model was used to predict the dyeing formula of 8 groups of standard samples, and the experimental results were evaluated by the CIEDE2000 color difference evaluation standard. The experimental results show that the ΔE₀₀ between the fitted sample and the standard sample obtained by predicting the dyeing formula based on the ADE algorithm optimization model is less than 3. The number of iterations is 13.04% of the genetic algorithm and 50.00% of the differential evolution, and the optimization speed is faster. For standard sample 1, the new formula predicts that the ΔE₀₀ between the fitted sample and the standard sample is about 1/5, 1/3, and 1/2 of the least squares, GA and DE algorithm optimization models, the prediction accuracy is higher, which shows that the new model has higher value in the full spectrum color matching of dyed wood veneer.