تحلیلی بر پراکنش مراکز بیمارستانی با رویکرد پدافند غیرعامل با هدف مدیریت مخاطرات با استفاده از شبکۀ عصبی مصنوعی (مطالعۀ موردی: شهر تبریز)

بیمارستان از اجزای حیاتی و حساس شهر است که از نظر شرایط مربوط به پدافند غیرعامل به‌منظور مدیریت مخاطرات باید توجه ویژه‌ای به آن شود. شهر تبریز یکی از شهرهای پرجمعیت کشور است که خدمات درمانی مورد نیاز کل استان آذربایجان شرقی و تا حدودی استان‌های همجوار را تامین می‌کند. از این‌رو باید برای مقابله با مخاطرات طبیعی و انسانی رویکردهای لازم صورت گیرد تا بیمارستان‌های شهر در مواجهه با خطرهای احتمالی حداکثر ایمنی را داشته باشند. به‌دلیل اهمیت مطالعۀ حاضر از 13 معیار اثرگذار فاصله از مراکز حساس نظامی و صنعتی، فاصله از ایستگاه‌های آتش‌نشانی، فاصله از پارک‌ها و فضای سبز، فاصله از مراکز بهداشتی درمانی، فاصله از مراکز تجاری، پراکندگی و شعاع دسترسی بیمارستان‌ها، فاصله از معابر و راه‌های اصلی، فاصله از گسل، فاصله از رودخانه، فاصله از کاربری‌های آموزشی، فاصله از مراکز سوخت و تراکم جمعیت استفاده شد و با روش شبکۀ عصبی موقعیت بیمارستان‌های شهر تبریز بررسی شد. نو‌آوری این پژوهش، استفاده از روش شبکۀ عصبی در تعیین آسیب‌پذیری مناطق مختلف در برابر مخاطرات و تهدیدهاست. برای این منظور از روش شبکۀ عصبی پرسپترون چندلایه (mlp) با الگوریتم آموزش لونبرگمارکوات استفاده شد. برای رسیدن به اهداف پژوهش 104 نقطۀ آموزشی معرفی و 13 نورون به‌عنوان نورون‌های لایۀ میانی تعیین شد و برای رسیدن به این تعداد از روش آزمون و خطا استفاده شد. با تعیین نحوۀ اثرگذاری معیارها مناطق مناسب مشخص شد. نتایج تحقیق نشان‌دهندۀ این است که در بین بیمارستان‌های کنونی بیمارستان شهید عالی‌نسب بهترین موقعیت را از نظر پدافند غیرعامل با هدف مدیریت مخاطرات در بین بیمارستان‌های شهر تبریز دارد و بیمارستان‌های امام رضا، شهید مدنی و طالقانی وضعیت نامساعدتری نسبت به بقیه دارند. نتایج همچنین نشان‌دهندۀ آن است که از نظر پدافند غیرعامل در مدیریت مخاطرات، قسمت‌های جنوب شرقی شهر نامناسب‌ترین و قسمت‌های شمال غربی بهترین مناطق برای احداث بیمارستان جدید هستند.

An Analysis of the Distribution of Hospital Centers with Passive Defense Approach to Hazard Management using Neural Network (Case Study: Tabriz City)

IntroductionHospital as one of the critical parts of the city should be considered in terms of passive defense to hazard management. Since the city of Tabriz has a large supply of medical services, it is necessary that the hospitals in this city be investigated and suitable parts from the perspective of passive defense are presented. In the present study, using 13 effective criteria and using neural network method, we investigated the position of hospitals in Tabriz city. The criteria include: distance from fire stations, distance from industrial and military centers, distance from parks and urban green spaces, distance from health centers, distance from main roads, population density, distance from faults, distance from riverside, Distance from training centers, distance from business premises, distance from warehouse and urban facilities, distance from fuel centers and dispersal and access radius of hospitals. The layers were first straightened and standardized in the Arc map, then using Neural Networks method using MATLAB software. To achieve the research goals, 104 educational points were introduced to the system. The results of the study indicate the status of hospitals in terms of nonoperating defense that among the current hospitals of ShahidAliNasab had the best position in terms of passive defense to hazard management, and the hospitals of Imam Reza, Madani and Taleghani had a more unfavorable situation than others, as well as the northwest of Tabriz city. The best conditions regarding the passive defense and hazard management features are northwest parts of city for building new hospitals.MethodologyTo work with Artificial Neural Network, firstly, effective parameters to locate the hospital should be provided to the network as input layers. Then, a number of training points should be given to the network so that the network uses the training points to determine the impact of each of the input layers; in fact, the network has received the training needed to cope with new areas. Finally, the entire city of Tabriz has been provided to the trained network, and in the end, the network, using what has been learned, identifies the optimal locations for the hospital.The neural network has several types. In this study, Multilayer Perceptron method was used with backpropagation algorithm to determine the optimal location of the hospital.Result and discussionAfter providing the necessary trainings to the network and run it, the result of the map of the optimal locations for the construction of the hospital was presented. The output is a valuable layer between zero and one.ConclusionThe city of Tabriz is the provider of many health services and the city as one of the big cities of Iran has the potential to attack enemy targets during a possible attack. On the other hand, there are natural risk factors in the city. The need to examine the status of the current hospitals finding the right place to build a hospital. According to the mentioned cases, necessary studies were carried out and 13 criteria for the mentioned goals were determined. These criteria were determined using previous studies and natural conditions of the city.The method used in this study was a neural network method, which was used with multilayer perceptron method with error propagation algorithm. We have plotted results with a numerical value from zero to one. By examining the position of the current hospitals in Tabriz, it seems hospitals are located in the central regions of the city and are more likely to follow the population factor.The results show that Ali Nasab Hospital has a better position in terms of passive defense to hazard management than other hospitals. The hospital also has an appropriate distance from the environmental threats, such as faults and streams, and in the distance from the critical centers at the time of target of a possible enemy attack, and it has access to the main routes and the first arteries, and open green spaces, which, in times of crisis, provide space for emergency relief and emergency hospital establishment. After that, Amir AlMomenin and Artesh hospital have a more favorable situation than other hospitals. Shahid madani, Imam Reza, and Taleghani hospitals have the worst conditions than other hospitals. Considering that, most hospitals are in Tabriz 1 and 2 districts and those have no radial distances with other hospitals in the areas that two Shahid Madani and Imam Reza hospitals are these centers. Because the hospital should be have distance at least 1000 meters from the fuel supply centers, this is not adhered to the case in the Taleghani Hospital. There are two fuel centers in the 1000meter hospital privacy. Accordingly, the southeast parts of Tabriz is the most unsuitable areas for construction of new hospitals, and the northwest were identified as the best areas for building new hospitals.