This paper presents unsupervised holding rule generation for a climbing robot Bernoulli holding pad, based on Evolutionary Algorithms. Dynamic variations in frictional coefficient of surface and robot state require adaptability on holding force, when combined with internal pad dynamics. A hybrid Evolutionary Algorithm (hEA) combining operators from Differential Evolution, Memetic Algorithm and Multi Objective aNd open-Ended Evolution (MONEE) Implementation was developed, tested and validated for augmenting robot pad dynamics in a dynamic environment, with the aim of reduction in pressure energy usage. Different variable combinations established in literature were permuted and results observed in terms of solution standard and speed. Comparative results showed the hEA method produced better results than other algorithms in terms of solution quality and processing time. Hence the work exhibited that hEA can improve holding force adaptability by providing quickly a set of near optimum conditions for the climbing robot.