Abstract: Automated and intelligent light supplement systems have been widely applied in greenhouses in recent years. Among them, the light environment regulation model can be the core content of the system. However, the existing models cannot consider the comprehensive influence of light quality and light intensity, as well as the double optimization of net photosynthetic rate and light use efficiency. In this study, a collaborative control method was proposed for the light quality and light intensity using multi-objective optimization, particularly for the efficient supplemental illumination of cucumbers in greenhouses. Firstly, a multi-factor coupled photosynthetic experiment was designed to obtain the net photosynthetic rate of cucumber leaves. The model of net photosynthetic rate was then established using support vector regression with temperature, carbon dioxide concentration, photosynthetic photon flux density and light quality ratio as the input, while the net photosynthetic rate as the output. Furthermore, the light use efficiency was calculated at the leaf scale, according to the definition. Secondly, a multi-objective optimization model was constructed with the light use efficiency and net photosynthetic rate as optimization targets, while the light quality and light intensity as control variables. The non-inferior solution set was solved using the multi-objective particle swarm optimization. Technique for order preference by similarity to ideal solution was used to select the control single point of light quality and light intensity, in order to narrow the regulation interval for the less subjectivity of manual selection. Finally, the red and blue light demand were calculated according to the multiple relationship of light quality and light intensity. And then the red and blue light models were fitted by support vector regression with the temperature and carbon dioxide concentration as the input. The control experiments were carried out to compare with the fixed light quality supplement and the photosynthetic maximum supplement, in order to verify the superiority. The theoretical verification experiment showed that the net photosynthetic rate decreased by 21.39%, whereas, the light demand decreased by 59.40%, compared with the photosynthetic maximum supplement. The net photosynthetic rate increased by 3.66% and 9.69%, respectively, compared with the fixed light quality of 0.5 and 0.8. The practical verification experiment was also carried out to further verify the energy efficiency. The results showed that the physiological indicators were better than the fixed light quality supplement under similar power consumption, indicating significant differences in the stem diameter, dry weight and strong seedling index. There was no significant difference in physiological indicators, but the power consumption decreased by 27.43%, compared with the photosynthetic maximum supplement. The consumption of light and electrical energy resources was effectively saved to keep the physiological indicators almost unchanged. The model construction can be expected to serve as the new perspective for the greenhouse light supplement. This study can provide a new light supplement strategy for the facility's agricultural regulation and the efficient utilization of agricultural production resources.