![anylogic 6.4.1 anylogic 6.4.1](https://www.anylogic.ru/upload/medialibrary/9bd/9bd78856c4e826d916454a57130d0a0c.jpg)
![anylogic 6.4.1 anylogic 6.4.1](https://img.informer.com/screenshots/6226/6226274_1.png)
The resulting tool is used by the ARSIF for both strategic and operational control of evacuation in the event of a major flood.
![anylogic 6.4.1 anylogic 6.4.1](http://www.downxia.com/uploadfiles/2020/0528/20200528094209170.jpg)
We consider hospital capacities on a macroscopic level: our model does not include detailed evacuation decisions from each building as proposed in several papers , but it allows to control the flow of patients over the different facilities. Hospital evacuation decisions adapt to the evolution of the flood and the model is used for both preparedness before the event and response during the event. Unlike most of the articles in literature, we integrate in our model (i) the flood dynamics (movement in space and time) instead of considering the impact at a predefined time using a Markov model, and (ii) the macroscopic patient flow model on a long horizon (one year) using a discrete-event model. The main objective of this study consists in proposing a management tool for hospital evacuation planning in the event of a major flood of nearby rivers. A practical study case that is the resilience of the healthcare facilities network of Île-de-France region (Paris, France) in case of a major flood is studied in this paper. In particular, ARSIF is working and the anticipation, preparation and management of health crises. The mission of ARSIF is to implement health policy in the region. This research is supported by the Île-de-France Regional Health Agency ( Agence Régionale de Santé d’Île-de-France, ARSIF).
In addition, when a hospital is flooded, its evacuation is more constrained than mass evacuation due to the patients’ health conditions and the necessity to relocate them in appropriate facilities. However, hospitals may be themselves at risk of flood and therefore there is a simultaneous (i) increase of the demand for emergency services and (ii) decrease of their provision. Health facilities represent a key resource and must accommodate the usual patients flow and the flow resulting from a disaster . One of the main challenges in DOM is the lack of resources , and understanding and managing critical resources is essential for the minimization of the disaster impact. A classified survey of the use of OR techniques for DOM is presented in and updated in . Artificial neural network is also used in as well as geographical information system to predict the dynamic of water during the flood and to optimize the gate-control strategies. Wei et al. use artificial neural network to predict flood occurrence and prepare for the impact. Lian and Yen compares several risk calculation methods applied to flood mitigation. Operations Research (OR) techniques can be used in several components of the DOM cycle. Recovery includes actions taken to return to a normal situation after a disaster. A response is the application of the preparedness plans during the disaster by allocating the necessary resources to protect the community. Preparedness activities take place before the disaster occurs to prepare the right response. Mitigation activities take place before and after emergencies and aim at reducing the occurrence and the impact of a disaster. The Federal Emergency Management Agency defines DOM as a cycle of four components: mitigation, preparedness, response and recovery (Fig. 1). Disaster Operations Management (DOM) is a set of actions and decisions aimed at preparing responses to natural or man-made disasters by reallocating resources (e.g. health facilities, transportation). Like the majority of disasters, natural or man-made, floods can result in significant economic loss and human casualties.