基于LSSVM和NSGA-Ⅱ混凝土耐久性多目标配合比优化

为减少由于耐久性不足而导致的混凝土结构劣化损伤，实现快速准确地确定配合比优化方案，以吉林某隧道工程为背景，选择混凝土耐久性主要指标抗冻性（动弹性模量）和抗渗性（氯离子渗透系数）为研究目标，建立一种LSSVM-NSGA-Ⅱ算法。基于原材料及配合比，利用最小二乘支持向量机（LSSVM）模型实现混凝土动弹性模量和氯离子渗透系数耐久性指标高精度预测，将回归预测函数作为适应度函数并结合规范及工程项目要求建立原材料及配合比相关的约束条件；在此基础上，利用NSGA-Ⅱ实现多目标优化。研究结果表明： 1）将相对动弹性模量和氯离子渗透系数作为耐久性评价指标并利用LSSVM模型进行预测的结果精度高； 2）所得预测回归函数作为适应度函数，以混凝土抗冻性和抗渗性为目标，利用NSGA-Ⅱ算法进行多目标优化后，可以获得配合比优化目标值； 3）经试验可知，混凝土工作性能等均符合规范和工程项目要求，且与工程实际情况相符。研究结果反映了该模型在配合比多目标寻优中的智能化、精准化。

Optimization of Multi Objective Mix Ratio for Concrete Durability Basedon LSSVM and NSGA Ⅱ

The concrete structure is often degraded and damaged due to insufficient durability. In order to realize rapid and accurate determination of the optimization plan for concrete mix ratio, an LSSVM NSGA Ⅱ algorithm is established by selecting the frost resistance(dynamic elastic modulus) and anti permeability(chloride ion permeability coefficient) of main indices of concrete durability of a tunnel project in Jilin as study targets. The high precision prediction of the concrete dynamic elastic modulus and durability index of chloride ion permeability coefficient is realized by using the least squares support vector machine (LSSVM) model based on raw materials and mix ratio. The regression prediction function is used as the fitness function and the constraint conditions of raw materials and mix ratio are established considering specifications and project requirements, and based on which the multi objective optimization is achieved by NSGA Ⅱ. The results show that: (1) The prediction precision of the method mentioned above is high. (2) The targeted optimization values of concrete mix ratio can be obtained by regression function obtained and NSGA Ⅱ algorithm. (3) The performance of concrete prepared can meet relevant specifications and project requirements, and conforms to the actual conditions of the project. The results indicate the intelligence and precision of the model in the multi objective optimization of concrete mix ratio.