Research on modeling process and optimization algorithm of three-dimensional collaborative design of pumped storage power station based on BIM and rule engine

In view of the challenges of complex topography and geology, multi-discipline intersection and long design cycle in the construction of pumped storage power station, it is difficult for the traditional two-dimensional design mode to deal with them efficiently. This study proposes a three-dimensional collaborative design framework and optimization algorithm based on BIM and rule engine. In the modeling process, firstly, the coordinate system and format of multi-source data are unified through data standardization and basic modeling, and a high-precision basic information model is constructed, including generating a three-dimensional geological model by using UAV aerial survey, geological radar and other data, and realizing the accurate alignment of multi-professional models through coordinate transformation; Subsequently, in the collaborative development stage of professional models, each major carries out interactive modeling based on a unified platform to realize dynamic data transmission and two-way collaboration, such as information interaction between hydraulic and electromechanical majors, and at the same time constructs a "three-level conflict detection rule" system to realize dynamic coordination of conflicts from geometric collision, professional specifications to construction feasibility. In the aspect of optimization algorithm, the rule engine architecture with Drools as the core is constructed, and the design specification is transformed into executable rules to realize the automatic verification of design parameters. Aiming at the problem of spatial layout, an improved particle swarm optimization (PSO) is proposed to optimize the equipment layout through multi-objective fitness function and conflict-oriented speed correction mechanism. For key hydraulic structures, NSGA-II is used for multi-objective parameter optimization to balance investment and operation benefits. The case study shows that this method shortens the design cycle by 53%, improves the equipment space utilization rate by 15.2%, reduces the steel consumption by 2,800 tons, reduces the design rework cost by 85.5%, and indirectly reduces the fossil energy consumption by about 500,000 tons of standard coal every year, which provides theoretical and practical support for the intelligent design of pumped storage power stations, and helps to improve the flexibility of power systems and achieve the goal of "double carbon".