Cost prediction for construction projects provides important information for project feasibility studies and design scheme selection. To improve the accuracy of early-stage cost estimation for construction projects, an improved neural network prediction model was proposed based on BP (back propagation) neural network and Snake Optimizer algorithm (SO). SO algorithm is adopted to optimize the initial weights and thresholds of the BP neural network. Cost data for 50 construction projects undertaken by Shandong Tianqi Real Estate Group in China was collected, and the data samples were clustered into three categories using cluster analysis. 18 engineering feature indicators were determined through a literature review and 10 feature indicators were selected using Boruta algorithm for the input set. Compared to BP neural network and PSO–BP neural network, the results show that the improved SO–BP model has higher prediction accuracy, stability, better generalization ability and applicability. Therefore, based on reasonable feature indicators, the method proposed in this paper has certain guiding significance for predicting engineering costs.

This paper proposes a method to optimize reinforcement layout of three-dimensional members under a state of complex stress and multiple load cases (MLCs). To simulate three-dimensional members, the spatial truss-like material model is adopted. Three families of truss-like members along orthotropic directions are embedded continuously in concrete. The optimal reinforcement layout design is obtained by optimizing the member densities and orientations. The optimal design of three-dimensional member is carried out by solving the problem of minimum volume of reinforcing bars with stress constraints. Firstly, the optimized reinforcement layout under each single load case (SLC) is obtained as per the fully stressed criterion. Second, on the basis of the previous results, an equivalent multi-case optimization is proposed by introducing the idea of stiffness envelope. Finally, according to the characteristics of the truss-like material, a closed and symmetrical surface is adopted to fit the maximum directional stiffness under all SLCs. It can be proved that the densities and orientations of truss-like members are the eigenvalues and eigenvectors of the surface coefficient matrix, respectively. Several three-dimensional members are used as examples to demonstrate the capability of the proposed method in finding the best reinforcement layout design of each reinforced concrete (RC) member and to verify its efficiency in application to real design problems.

The paper proposes a procedure for the conceptual design of reinforced concrete (RC) structures under a multiple load case (MLC), based on the truss-like topology optimization method. It is assumed that planar truss-like members are densely embedded in concrete to simulate RC structures. The densities and orientations of the reinforcing bars at nodes are regarded as optimization variables. The optimal reinforcement layout is obtained by solving the problem of minimizing the total volume of reinforcing bars with stress constraints. By solving a least squares problem, the optimized reinforcement layout under theMLCis obtained.According to the actual needs of the project, the zones to be reinforced are determined by reserving a certain percentage of elements. Lastly, a recommended reinforcement design is determined based on the densities and orientations of truss-like members. The reinforcement design tends to be more perfect by adding necessary structural reinforcements that meet specification requirements. No concrete cover is considered. Several examples are used to demonstrate the capability of the proposed method in finding the best reinforcement layout design.