
تعداد نشریات | 11 |
تعداد شمارهها | 224 |
تعداد مقالات | 2,246 |
تعداد مشاهده مقاله | 3,362,257 |
تعداد دریافت فایل اصل مقاله | 2,415,024 |
بهینه سازی چند هدفه جانمایی دوربین مدار بسته با استفاده از الگوریتم های هوش مصنوعی | ||
پژوهش های سنجش از دور و اطلاعات مکانی | ||
مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 15 شهریور 1404 | ||
نوع مقاله: مقاله پژوهشی | ||
شناسه دیجیتال (DOI): 10.22061/jrsgr.2025.12191.1103 | ||
نویسندگان | ||
حسن صانعی آرانی1؛ مهدی اسماعیلی* 2؛ محمد علی افشار کاظمی3 | ||
1گروه مدیریت فن آوری اطلاعات ، دانشکده مدیریت و اقتصاد، دانشگاه آزاد اسلامی- واحد علوم و تحقیقات، تهران، ایران | ||
2گروه مهندسی کامپیوتر دانشگاه آزاد اسلامی- واحد کاشان، ;کاشان، ایران | ||
3گروه مدیریت صنعتی ، دانشکده مدیریت و اقتصاد ، دانشگاه آزاد اسلامی- واحد تهران مرکز، تهران، ایران | ||
تاریخ دریافت: 08 اردیبهشت 1404، تاریخ بازنگری: 20 تیر 1404، تاریخ پذیرش: 01 شهریور 1404 | ||
چکیده | ||
بهینهسازی جانمایی دوربینهای مداربسته به عنوان یکی از ارکان اساسی سیستمهای هوشمند مدیریت ترافیک شهری است. روش این کار بر پایهی تحلیل نقشههای شهری ست و نیازمند یک نقشه جامع و دقیق از شهر است تا بتواند موقعیتهای بهینه را شناسایی کند. نقشه مورد استفاده را میتوان با یک ماتریس نمایش داد که بهصورت یک شبکه دوبعدی از نقاط است و مسیرهای قابل دسترس و موانع غیرقابل عبور با اعداد مختلف تعریف شدهاند.فرآیند جانمایی بهینه پس از تشکیل ماتریس مدل، در چهار مرحله به طور سیستماتیک انجام میشود. در مرحله اول، با استفاده از توزیع احتمال مبتنی بر تراکم جمعیتی ، جفتهای مبدأ-مقصد بهصورت تصادفی تولید میشوند. در مرحله دوم برای هر جفت مبدأ-مقصد گام قبل ، مسیریابی بهینه با شبیهسازی رفتار ترافیکی شهروندان، در دو رویکرد ساعات عادی و انتخاب کوتاهترین مسیر و ساعت شلوغی با انتخاب مسیرهای فرعی صورت می گیرد. در مرحله سوم با تجمیع تمام مسیرهای تولیدشده، ترافیک مجازی ساخته میشود. سپس سوم تراکم مسیرها محاسبه و بهینه سازی بر اساس ترافیک انجام می شود. در مرحله چهارم، با در نظر گرفتن انواع دوربین بر اساس قیمت خرید و هزینههای نصب، جانمایی بر اساس هزینه بهینه میگردد.یک صد هزار داده جدید ایجاد شد و سپس دو آزمایش ترتیب داده شد. در آزمایش اول، از یک الگوریتم حریصانه برای ماکزیمم کردن پوشش دوربینها و آزمایش دوم از روش پیشنهادی استفاده می کند. نتایج نشان داد روش پیشنهادی در پایش مسیرهای جدید 40 درصد کاراتر و انجام پروژه 6/6 درصد مقرون به صرفه تر است. | ||
عنوان مقاله [English] | ||
Multi-objective optimization of CCTV camera placement using artificial intelligence algorithms | ||
نویسندگان [English] | ||
hassan sanei arani1؛ mahdi esmaeili2؛ mohamad ali afshar kazimi3 | ||
1Department of Information Technology Management , Faculty of Management and Economics, Islamic Azad University- Science and Research branch, Tehran, Iran | ||
2Department of Computer Engineering,, Islamic Azad University- Kashan Branch, Kashan, Iran | ||
3Department of Industrial Management , Faculty of Management and Accounting, Islamic Azad University- Central Tehran branch,Tehran, Iran | ||
چکیده [English] | ||
Background and Objectives: Optimization of CCTV camera placement is one of the fundamental pillars of smart urban traffic management systems. The correct deployment of these cameras can greatly affect the accuracy of traffic monitoring and reduce the time to detect incidents, so the problem of optimizing camera placement has been a research challenge for many researchers for many years. Modern approaches to solving the problem are based on multi-objective optimization methods to enable simultaneous analysis of different and effective parameters. Despite significant advances in optimization methods, current approaches are based on two-dimensional and three-dimensional gridding of the studied space, which face fundamental limitations in complex urban environments. In these methods, the desired space is divided into a regular grid and the optimal points for installing cameras are selected with appropriate angular rotation. However, in the real topology of cities, the road networks are spread out as nested and irregular lines, so that many of the calculated points are located outside the accessible routes. This discrepancy between theoretical models and practical conditions seriously questions the effectiveness of traditional methods. Given these limitations, it has become a necessity to provide a new framework that can simultaneously consider the real topology of cities, physical constraints and urban planning requirements. New methods must be able to integrate real traffic routes, permitted camera installation locations and mandatory angles into their models. This requires the use of methods based on realistic virtual traffic data and artificial intelligence algorithms for optimization. Methods: The method is based on the analysis of urban maps and requires a comprehensive and accurate map of the city to identify optimal locations. The map used can be represented by a matrix that is a two-dimensional network of points, and accessible routes and impassable obstacles are defined by different numbers. Since the width of a street consists of several points, a row from the center of the route is considered as a representative of the route to restrict the movement of vehicles to pass through it and provide an ideal location for the deployment of surveillance cameras. The optimal placement process is systematically carried out in four stages after the formation of the model matrix. In the first stage, origin-destination pairs are generated randomly using a probability distribution based on population density. In the second stage, for each origin-destination pair of the previous step, optimal routing is performed by simulating the traffic behavior of citizens in two approaches: normal hours and selecting the shortest route, and rush hour by selecting secondary routes. In the third stage, virtual traffic is created by aggregating all the generated routes. Then, the density of the routes is calculated and optimization is performed based on traffic. In the fourth stage, by considering the types of cameras based on purchase price and installation costs, placement is optimized based on cost. Findings: One hundred thousand new data were created and then two experiments were conducted. In the first experiment, a greedy algorithm was used to maximize the coverage of the cameras along the entire route. The second experiment uses the proposed method. First, high-traffic spots are identified based on the aforementioned relationships, and then, with the aim of maximizing the coverage in these high-traffic spots and then minimizing the installation costs, the cameras are installed in optimal locations. The results showed that the proposed method is 40% more efficient in monitoring new routes and 6.6% more cost-effective in carrying out the project. Conclusion: In the placement of urban CCTV cameras, any method that considers maximum coverage on urban routes will not be effective, and traffic measurement is an important factor for optimization. Also, in the proposed method, since the geometric features of the routes are eliminated, this method is scalable and can be applied to any city and routing system. Also, urban planners usually purchase cameras with different fields of view and brands, which can be considered as an opportunity and a secondary goal of optimization to reduce costs. | ||
مراجع | ||
| ||
آمار تعداد مشاهده مقاله: 2 |