The existing energy network security defense system uses firewall, intrusion detection, host monitoring, identity authentication, anti-virus software and vulnerability repair to build a fortress type rigid defense system to block or isolate external invasion. This static layered deep defense system is based on prior knowledge and has the advantages of rapid response and effective protection in the face of constant attacks When confronting the unknown attacking opponent, he is not able to do his best, and he is in danger of being attacked easily. In this context, multi-objective decision has more than two decision-making objectives and needs to use multiple criteria to evaluate and optimize the decision-making of alternatives. Due to the objectives of economic benefit, safety in production and environmental protection, it is necessary to use a variety of criteria to evaluate and optimize schemes. In this paper, we propose RBF neural network and weight-based algorithm to achieve multi-objective decision. We leverage RBF neural network to construct objective weight assignment model. The goal of our weight-based algorithm is that the multi-objective optimization problem is formulated as a single-objective optimization problem by assigning certain weights to each objective, and then the non-inferior solution of the multi-objective optimization problem is generated by changing the weights of each objective extensive.

现有的能源网络安全防御系统使用防火墙，入侵检测，主机监视，身份认证，防病毒软件和漏洞修复来构建要塞类型的刚性防御系统，以阻止或隔离外部入侵。这种静态的分层深度防御系统是基于先验知识的，在持续不断的攻击面前具有快速反应和有效保护的优势。当面对未知的攻击对手时，他无法尽力而为，并且有被攻击的危险。容易受到攻击。在这种情况下，多目标决策具有两个以上的决策目标，并且需要使用多个标准来评估和优化替代方案的决策。出于经济利益，生产安全和环境保护的目标，有必要使用各种标准来评估和优化方案。在本文中，我们提出了RBF神经网络和基于权重的算法来实现多目标决策。我们利用RBF神经网络构建客观的权重分配模型。我们基于权重的算法的目标是，通过为每个目标分配一定的权重，将多目标优化问题表述为单目标优化问题，然后通过以下方法生成多目标优化问题的非劣解：广泛地改变每个目标的权重。