كليدواژه :
منطق فازي , شبكه عصبي مصنوعي , gps , يون‌سپهر , tec , سامانه استنتاج فازي و ايران
چكيده فارسي :
در اين مقاله مقدار محتواي الكترون كلي (tec) لايه يون‌سپهر با استفاده از سامانه استنتاج فازي (fis) مدل‌سازي شده است. نوآوري اصلي اين پژوهش، مدل‌سازي سري زماني تغييرات tec در ايران با استفاده از fis است. براي آموزش شبكه فازي، از الگوريتم آموزش بهينه سازي انبوه ذرات هيبريد (bp-pso) استفاده شده است. اين الگوريتم آموزش، در مراحل اوليه جستجوي جواب از الگوريتم بهينه سازي انبوه ذرات (pso) و در نزديكي جواب بهينه از الگوريتم پس انتشار خطا (bp) بهره مي برد. از مشاهدات سال 2015 ايستگاه gps تهران، كه يكي از ايستگاه هاي شبكه جهاني igs است، جهت ارزيابي مدل پيشنهادي استفاده شده و نتايج كار با نتايج خروجي هاي شبكه جهاني igs (gim-tec) و مدل شبكه عصبي مصنوعي (ann) با ساختار 2-18-1 مقايسه شده است. جهت ارزيابي دقت و صحت مدل شبكه فازي ارائه شده در اين مقاله، از هر فصل، پنج روز براي داده آزمون انتخاب شده و اعتبارسنجي مدل در اين بيست روز صورت گرفته است. براساس نتايج، ميانگين خطاي نسبي محاسبه شده در بيست روز مورد آزمون براي مدل fis، ann و gim در مقايسه با gps به ترتيب برابر با 11.25%، 19.68% و 16.03% است. همچنين ميانگين خطاي مطلق محاسبه شده براي مدل fis، ann و gim در مقايسه با gps در بيست روز مورد آزمون به ترتيب برابر با tecu 1.32، tecu 3.33 و tecu1.98 است. ضريب همبستگي محاسبه شده در بيست روز مورد آزمون براي fis، ann و gim در مقايسه با gps به ترتيب برابر با 0.9474، 0.6960 و 0.831 به دست آمده است. موقعيت ايستگاه gps تهران براساس tec حاصل از سه مدل fis، ann و gim و با استفاده از تعيين موقعيت نقطه‌اي دقيق (ppp) محاسبه شده است؛. طبق اين محاسبه، مدل شبكه فازي نسبت به دو مدل ديگر، خطاي كمتري در تعيين موقعيت ايستگاه تهران دارد. نتايج تحليل ها حاكي از برتري مدل fis در مقايسه با مدل ann و gim است. با استفاده از مدل پيشنهادي اين پژوهش مي توان سري زماني محتواي الكتروني كلي يون‌سپهر را با دقت و صحت زياد مدل‌سازي و بررسي كرد. اين مدل مي تواند جايگزيني مناسب براي خروجي هاي شبكه جهاني igs در منطقه ايران باشد.
چكيده لاتين :
Ionosphere is a layer in the upper part of the atmosphere wide-ranging from 60 km to 2000 km.
It has a very significant role in radio wave propagation because of its electromagnetic
attributes. Ionosphere is mainly affected by solar zenith angle and solar activity. In the daytime,
ionization in ionosphere is at the highest level and the ionospheric effects are stronger. In
the night-time, ionization decreases and the effects of ionosphere gets weaker. One of the most
important parameters that defines the physical structure of ionosphere is Total Electron Content
(TEC). TEC is a line integral of electron density along signal path between satellite to the
receiver on the ground. The unit of TEC is TECU and 1 TECU equals 1016 electrons/m2. The
TEC values can be computed from dual frequency Global Positioning System (GPS) stations,
which are the most available observations for studying the Earth’s ionosphere. However,
because of scattered repartition of dual frequency of GPS stations, precise information on TEC
over the favorable region is unknown.
Fuzzy inference systems (FIS) take inputs and process them based on the pre-specified rules
to produce the outputs. Both the inputs and outputs are real values, whereas the internal
processing is based on fuzzy rules and fuzzy arithmetic. FIS is the key unit of a fuzzy logic
system having decision making as its primary work. It uses the “IF…THEN” rules along with
connectors “OR” or “AND” for drawing essential decision rules. A FIS is defined according to
the following five main sections :
• Rule Base − It contains fuzzy IF-THEN rules;
• Database − It defines the membership functions of fuzzy sets used in fuzzy rules;
• Decision-making Unit − It performs operation on rules;
• Fuzzification Interface Unit − It converts the crisp quantities into fuzzy quantities; and
• Defuzzification Interface Unit − It converts the fuzzy quantities into crisp quantities .
In this paper, the TEC of the ionosphere is modeled using FIS. The fuzzy inference system
uses the rules IF-THEN to recognize the characteristics of dynamic phenomena. This feature,
along with the simplicity of computing, has made it possible for this model to study the
temporal and spatial variations of the ionosphere. In fact, the main innovation of the paper is
the time series modeling of TEC in Iran using FIS. Hybrid particle swarm optimization training
(BP-PSO) algorithm is used to train fuzzy network. This algorithm uses the PSO in the early
stages of searching for solution and uses the back propagation (BP) near the optimal solution.
From the observations of 2015, the Tehran GPS station, which is one of the IGS global
stations, was used for evaluation of the proposed model. Also, the results were compared with
the results of the global ionosphere map (GIM) TEC as well as artificial neural network model
(ANN). In order to evaluate the accuracy of the fuzzy model presented in this paper, 5 days of
each season were selected as the test data and model validation was performed in these 20
days. Based on the results, the average relative error calculated in the 20 test days for FIS,ANN and GIM models compared to GPS were 11.25%, 19.68% and 16.03%, respectively.
Besides, the average absolute error calculated for FIS, ANN and GIM models compared to
GPS in the 20 test days was 1.32 TECU, 3.33 TECU and 1.98 TECU, respectively. The
calculated correlation coefficients between TEC obtained from FIS, ANN and GIM compared
to GPS were 0.9474, 0.6960 and 0.831, respectively. The results of the analysis show that the
FIS model is superior to the ANN and GIM models. Using the proposed model of this research,
the time series of the ionosphere TEC can be modeled and investigated with high accuracy.
This model can also be a good alternative to the outputs of the IGS network in Iran.
