Particle filtering is a contemporary Sequential Monte Carlo state inference and identification methodology that allows filtering of general non-Gaussian and non-linear models in light of time series of empirical observations. Several previous lines of research have demonstrated the capacity to effectively apply particle filtering to low-dimensional compartmental transmission models. We demonstrate here implementation and evaluation of particle filtering to more complex compartmental transmission models for pertussis – including application with models involving 1, 2, and 32 age groups and with two distinct functional forms for contact matrices – using over 35 years of monthly and annual pre-vaccination provincial data from the mid-western Canadian province. Following evaluation of the predictive accuracy of these four particle filtering models, we then performed prediction, intervention experiments and outbreak classification analysis based on the most accurate model. Using that model, we contribute the first full-paper description of particle filter-informed intervention evaluation in health. We conclude that applying particle filtering with relatively high-dimensional pertussis transmission models, and incorporating time series of reported counts, can serve as a valuable technique to assist public health authorities in predicting pertussis outbreak evolution and classify whether there will be an outbreak or not in the next month (Area under the ROC Curve of 0.9) in the context of even aggregate monthly incoming empirical data. Within this use, the particle filtering models can moreover perform counterfactual analysis of interventions to assist the public health authorities in intervention planning. With its grounding in an understanding of disease mechanisms and a representation of the latent state of the system, when compared with other emerging applications of artificial intelligence techniques in outbreak projection, this technique further offers the advantages of high explanatory value and support for investigation of counterfactual scenarios.