A hybrid experimental learning method is applied to simulate the carbon-electricity market, which solves the macro-scale problems from the conflict between electricity market price reduction and carbon emission cost, and the micro-scale problems from the different decision time scales.