食品、饮水的重金属污染已经成为目前人类健康的一大威胁，很多重金属不是单独存在，而是两种或者多种以上同时存在，多种重金属在人或动物的体内会相互作用，产生与单一重金属严重程度不同的毒性。而且，在计算人类健康风险的过程中，往往由动物毒理实验结果依据体重或采用安全系数法进行简单外插，方法较为粗糙，基于生理药代动力学模型（PBPK模型）的外插方法更为精确科学。本文主要通过建立单一重金属的PBPK模型，模拟得到实验动物重金属暴露的内剂量；通过比较单一重金属PBPK模型的内剂量模拟值与复合暴露毒性实验各脏器重金属浓度实测值，探讨复合暴露情况下各重金属在体内吸收、分布及代谢过程的相互影响；通过比较复合重金属暴露和单一重金属暴露产生的毒性效应，探讨重金属复合暴露的联合毒性作用，为进一步进行更为精确的重金属复合暴露的人类健康风险评估奠定基础。本文主要分为三部分：第一部分，根据砷、镉、铅、铬在体内不同的代谢过程，利用Matlab/Simulink建立生理药代动力学模型，并进行调试，以使模型能够应用于这四种重金属在大鼠体内的代谢、分布过程。在利用一套实验数据进行参数的优化之后，用建立的模型和优化好的参数，输入另外的毒理实验数据，验证模型是否能够较好的模拟其他实验的数据。结果表明，模拟结果基本与文献中的实验结果一致。第二部分，将含有砷、镉、铬、铅、汞的混合溶液对大鼠进行染毒。分别比较了高、中、低剂量组与对照组相比大鼠体重的变化、测量了血液、肝脏、肾脏中各个重金属的含量，并且利用建立的四种重金属的PBPK模型进行模拟，与最终的实验结果进行比较，以研究在多个重金属同时存在的情况下，是否会影响彼此在组织器官中的分布。结果发现，砷的模拟结果小于实验值，镉、铅模拟结果基本与实验结果符合，铬模拟结果大于实验值。第三部分，将复合重金属暴露的毒理实验数据与相似文献中各重金属单一给药时的毒理实验数据进行比较，即观察在复合给药时，对大鼠体内各项指标效应的影响与单一给药时的差别。在比较中发现，单一重金属染毒比复合染毒时，砷、镉、铅、汞对大鼠产生的毒性效应更大，可能重金属之间产生了拮抗效应；铬单一染毒和复合染毒时对大鼠产生的毒性效应没有明显差异。

The pollution of heavy metals is a great threat of human health, the primary route of exposure for those metals is likely to be oral. Most of the time, those metals existed as mixtures, such as binary mixtures, trinary mixtures, and even more complex mixtures. When the mixtures enter human body, the joint toxicity would be different from that resulted from individual metals. And in the process of calculating human health risk assessment, extrapolation of results from animal experiment to humans’ is a routine, however, this method is not that accurate. This article focuses on the construction of PBPK model of individual metals. By using the models, simulate the concentration of certain metal in rat tissues (internal dose). By comparing the simulated internal doses and the experiment data, discuss the interaction of heavy metals’ ingestion, distribution and metabolism in the body; Bycomparing the differences of toxic effects resulted by joint metals and single metal, discuss the joint toxic effects of complex heavy metals. These works can be a further step on the way of calculating human health risk assessment accurately. This article has three main sections.First, according to the various metabolic processes of arsenic, cadmium, lead and chromium, construct four PBPK models in the Matlab/Simulink environment, respectively, which can simulate the metabolism, distribution and excretion of the four metals in rats. After the construction and parameter optimization of the models, use another dataset to evaluate the models. The results show that the simulated results are agreed with that in relative articles.Second, Forty male SD rats were all divided into four groups (control, low-dose, mid-dose and high-dose) of five animals each, and so were the forty female ones. These rats exposed to arsenic, cadmium, lead, chromium and mercury in drinking water. We recorded and compared the rats’body weight of the four groups, measured the heavy metals’ concentration in blood, liver and kidney and compared the data with the simulating results. By doing these, we can figure out if heavy metals have interactions with each other’s distribution in the targeted tissues when they are administrated at the same time.The results show that simulated arsenic concentration in liver and kidney is lower compared to experiment data, cadmium and lead’s simulated concentrations are agree with experiment results and simulated chromium tissue concentration is higher than experiment data.Third, we compared the joint toxic experiment results (mainly focus on the biomarkers and other related parameters) with that of former single metal toxic studies to determine whether joint toxic effect is more sever. It was surprised to find that individual arsenic, cadmium, lead and mercury may result in more severe injuries. The result indicates that there may existantagonizing affect among these metals. Toxic effects resulted by individual chromium and joint metals do not have significant differences.