•Mathematical and computational models are used to predict cases of COVID-19 in Mexico.•The data is obtained through the Daily Technical Report issued by the Mexican Ministry of Health.•Gompertz, ...Logistic and Artificial Neural Network perform the modeling of the cases confirmed by COVID-19 with an R2>0.999.•Logistic, Gompertz and inverse Artificial Neural Network predicts the maximum number of new daily cases on May 8th, June 25th and May12th, 2020, respectively.•The Gompertz, Logistic and inverse Artificial Neural Network models predict different number of cases of COVID-19 at the end of the epidemic.
This work presents the modeling and prediction of cases of COVID-19 infection in Mexico through mathematical and computational models using only the confirmed cases provided by the daily technical report COVID-19 MEXICO until May 8th. The mathematical models: Gompertz and Logistic, as well as the computational model: Artificial Neural Network were applied to carry out the modeling of the number of cases of COVID-19 infection from February 27th to May 8th. The results show a good fit between the observed data and those obtained by the Gompertz, Logistic and Artificial Neural Networks models with an R2 of 0.9998, 0.9996, 0.9999, respectively. The same mathematical models and inverse Artificial Neural Network were applied to predict the number of cases of COVID-19 infection from May 9th to 16th in order to analyze tendencies and extrapolate the projection until the end of the epidemic. The Gompertz model predicts a total of 47,576 cases, the Logistic model a total of 42,131 cases, and the inverse artificial neural network model a total of 44,245 as of May 16th. Finally, to predict the total number of COVID-19 infected until the end of the epidemic, the Gompertz, Logistic and inverse Artificial Neural Network model were used, predicting 469,917, 59,470 and 70,714 cases, respectively.
•ANNim-PSO and ANNim-GA methodology were implemented to improve the PTC performance.•Short computation time is required to optimize PTC's six input variables.•Rim angle, inlet temperature, and water ...flow were optimized simultaneously.•The rim angle was the most significant influence on PTC performance.•An optimal PTC thermal efficiency increase was achieved.
This work focused on presenting a multivariate inverse artificial neural network (ANNim) by developing two functions coupled to metaheuristic algorithms to increase a parabolic trough collector (PTC). This work aims to provide a new method capable of improving the thermal efficiency of a PTC by determining multiple optimal input variables. At first, two ANN models carried out to predict the PTC thermal efficiency (ηt), validated, and compared in detail. For that, six input parameters rim-angle (φr), inlet-temperature (Tin), ambient-temperature (Tamb), water volumetric flow rate (Fw), direct-solar-radiation (Gb) and wind-speed (Vv) considered as variables in the input layer. Two non-linear transfer functions (TANSIG and LOGSIG) in the hidden layer, a linear function (PURELIN) in the output layer, and the Levenberg-Marquardt training algorithm were applied. The results showed that both ANN models achieved satisfactory results with a coefficient of determination of 0.9511 and a root mean square error of 0.0193. Then, to get the variable's optimal values: rim-angle, inlet-temperature, and water volumetric flow rate, both ANN models inverted to acquire the multivariable objective function that could be resolved with genetic-algorithms (GA) and particle-swarm-optimization (PSO). The TANSIG function demonstrated better adaptation to the ANNim model by finding all the input variables in a random test with an error of 3.96% with a computational time of 14.39 s applying PSO. The results showed that by using the ANNim methodology, it is feasible to improve the performance of the PTC by optimizing from one, two, and three variables at the same time. In optimizing one variable at a time, it was possible to increase a random test's performance up to 54.78%, 27.62%, and 51.92% by finding the rim-angle inlet-temperature and water volumetric flow rate, respectively. In optimizing two variables simultaneously, it was possible to increase a random test's performance up to 36.73% by finding the appropriate inlet-temperature and water volumetric flow rate. In optimizing three variables simultaneously, it was possible to increase a random experimental test of up to 67.12%. Finally, the new ANNim method proposed may increase the thermal efficiency of a PTC in real-time because of the coupling of metaheuristic algorithms that allow obtaining optimal variables in the shortest possible time. Therefore, it can be a promising and widely used method for optimizing and controlling thermal processes.
In 2022, Mexico registered an increase in dengue cases compared to the previous year. On the other hand, the amount of precipitation reported annually was slightly less than the previous year. ...Similarly, the minimum-mean-maximum temperatures recorded annually were below the previous year. In the literature, it is possible to find studies focused on the spread of dengue only for some specific regions of Mexico. However, given the increase in the number of cases during 2022 in regions not considered by previously published works, this study covers cases reported in all states of the country. On the other hand, determining a relationship between the dynamics of dengue cases and climatic factors through a computational model can provide relevant information on the transmission of the virus. A multiple-learning computational approach was developed to simulate the number of the different risks of dengue cases according to the classification reported per epidemiological week by considering climatic factors in Mexico. For the development of the model, the data were obtained from the reports published in the Epidemiological Panorama of Dengue in Mexico and in the National Meteorological Service. The classification of non-severe dengue, dengue with warning signs, and severe dengue were modeled in parallel through an artificial neural network model. Five variables were considered to train the model: the monthly average of the minimum, mean, and maximum temperatures, the precipitation, and the number of the epidemiological week. The selection of variables in this work is focused on the spread of the different risks of dengue once the mosquito begins transmitting the virus. Therefore, temperature and precipitation were chosen as climatic factors due to the close relationship between the density of adult mosquitoes and the incidence of the disease. The Levenberg–Marquardt algorithm was applied to fit the coefficients during the learning process. In the results, the ANN model simulated the classification of the different risks of dengue with the following precisions (R
2
): 0.9684, 0.9721, and 0.8001 for non-severe dengue, with alarm signs and severe, respectively. Applying a correlation matrix and a sensitivity analysis of the ANN model coefficients, both the average minimum temperature and precipitation were relevant to predict the number of dengue cases. Finally, the information discovered in this work can support the decision-making of the Ministry of Health to avoid a syndemic between the increase in dengue cases and other seasonal diseases.
The present work is focused on modeling and predicting the cumulative number of deaths from COVID-19 in México by comparing an artificial neural network (ANN) with a Gompertz model applying multiple ...optimization algorithms for the estimation of coefficients and parameters, respectively. For the modeling process, the data published by the daily technical report COVID-19 in Mexico from March 19th to September 30th were used. The data published in the month of October were included to carry out the prediction. The results show a satisfactory comparison between the real data and those obtained by both models with a R
2
> 0.999. The Levenberg–Marquardt and BFGS quasi-Newton optimization algorithm were favorable for fitting the coefficients during learning in the ANN model due to their fast and precision, respectively. On the other hand, the Nelder–Mead simplex algorithm fitted the parameters of the Gompertz model faster by minimizing the sum of squares. Therefore, the ANN model better fits the real data using ten coefficients. However, the Gompertz model using three parameters converges in less computational time. In the prediction, the inverse ANN model was solved by a genetic algorithm obtaining the best precision with a maximum error of 2.22% per day, as opposed to the 5.48% of the Gompertz model with respect to the real data reported from November 1st to 15th. Finally, according to the coefficients and parameters obtained from both models with recent data, a total of 109,724 cumulative deaths for the inverse ANN model and 100,482 cumulative deaths for the Gompertz model were predicted for the end of 2020.
•Comparison of models with different mathematical approaches.•The prediction efficiency is not directly related to achieving the best fit.•Contrasting more than one predictive model to support ...decision-making.
This study was conducted to predict the number of COVID-19 cases, deaths and recoveries using reported data by the Algerian Ministry of health from February 25, 2020 to January 10, 2021. Four models were compared including Gompertz model, logistic model, Bertalanffy model and inverse artificial neural network (ANNi). Results showed that all the models showed a good fit between the predicted and the real data (R2 >0.97). In this study, we demonstrate that obtaining a good fit of real data is not directly related to a good prediction efficiency with future data. In predicting cases, the logistic model obtained the best precision with an error of 0.92% compared to the rest of the models studied. In deaths, the Gompertz model stood out with a minimum error of 1.14%. Finally, the ANNi model reached an error of 1.16% in the prediction of recovered cases in Algeria.
.
The Absorption Heat Transformer (AHT) has become an alternative proposal to weaken the trend of thermal pollution because it contributes to the recovery of residual heat and the use of solar thermal ...energy for its activation. Recently, compact designs have been developed that share two heat exchangers in one shell to reduce heat losses and save maintenance costs. The evaluation of the performance of AHTs depends on multiple parameters, which have been optimized using models based on ideal conditions. Multivariable optimization with artificial intelligence has considered experimental data to obtain feasible conditions related to the operation of the AHT. However, these models have focused only on determining one parameter at a time, reflecting a limitation. This study aims to develop a new optimization strategy to simultaneously maximize two relevant parameters associated with the efficiency of a compact experimental AHT. The optimization methodology is applied to increment the values of the heat generated in the absorber (QAB) and the Exergy Coefficient of Performance (ECOP) using the same objective function. This objective function is resolved when reaching the desired output parameters by determining multiple optimal conditions of the AHT. Initially, the modeling of the experimental data was carried out using an artificial neural network (ANN) for the diagnosis and prediction of the QAB and the ECOP. The model was validated by comparing the experimental values of QAB and ECOP against predicted values through a linear regression model, with a satisfactory result of R2>0.98. Subsequently, the multivariable inverse artificial neural network methodology for multiple output parameters (ANNim-m) was used and coupled with the Particle Swarm Optimization algorithm (ANNim-m-PSO) to generate the new multivariate optimization strategy and improve QAB and ECOP parameters simultaneously. Finally, 4 random tests with different initial operating conditions were optimized. Based on the optimization of the 4 tests, the QAB heat load was increased by 87.7 %, 54.2 %, 30.79 %, and 22.66 % from initial experimental conditions of 2.7, 9.0, 14.23 and 23.29 kW, respectively. In the case of ECOP, elevations of 28.5 %, 23.4 %, 15.4 %, and 16.18 % were obtained for initial values of 0.3801, 0.455, 0.5180, and 0.5729, respectively. It is determined that the parameters QAB and ECOP have a high sensitivity when the inlet temperature of the absorber of the external circuit (TAB) decreases. The results reveal that it is feasible to use the ANNim-m-PSO model, because the optimization was able to significantly maximize both QAB and ECOP parameters of the 40 kW experimental AHT.
•An experimental AHT of 40 kW is analyzed from the first and second laws of thermodynamics.•A new ANNim-m-PSO methodology is proposed for simultaneously improving the QAB and ECOP parameters of the AHT.•The optimization ANNim-m-PSO model was applied to 4 experimental randomized tests.•The QAB increased by 87.7, 54.2, 30.79, and 22.66 % from initial conditions of 2.7, 9.0, 14.23 and 23.29 kW, respectively.•The ECOP improved 28.5, 23.4, 15.4, and 16.18 % for initial values of 0.3801, 0.455, 0.5180, and 0.5729, respectively.
Display omitted
•ANNi is able to optimize multiple variables in AHT.•Multivariable function is solved by genetic algorithm.•Temperature in the generator is key variable to increase the COP.•The ...optimization of multiple variables is capable for energy savings.•Heat transformers can recycle part of useful energy into the heat supply.
This research presents an application of an artificial neural network inverse (ANNi) and genetic algorithm (GA) to propose and solve a multivariable function in order to optimize of an absorption heat transformer (AHT) with energy recycling. The purpose of the research is to provide a method capable of maximizing the coefficient of performance (COP) of the AHT, by finding multiple optimal input variables, with the benefit of minimizing energy in the heat supply of the equipment. AHT use waste heat sources to obtain useful energy, but by recycling part of the useful energy within the same system provides an increase in performance. Therefore, the research is focused on optimizing the heat source in the generator, evaporator and additionally the condenser simultaneously, since the value of the COP is determined through the ratio of the useful heat load between the loads of heat supplied. Consequently, this study is based on modeling the process and then performing the optimization, using the following research methods: artificial neural networks (ANN), ANNi and GA. An ANN model was developed to predict the value of COP, based on experimental data from the equipment. A satisfactory agreement was obtained by comparing the simulated and experimental data. With the ANN model consolidated, an ANNi was applied where the variables to be optimized were: temperature in the generator, temperature in the evaporator and temperature in the condenser. GA was chosen to solve the multivariable function. The results showed that when applying the ANNi-GA methodology, it is possible to carry out the multivariable optimization, since this methodology had only been used to optimize one variable at a time. The temperature in the generator turned out to be the key variable to increase the performance of AHT, followed by the evaporator and condenser, managing to maximize the COP value of a specific test from 0.26 to 0.43 and obtain an energy saving of up to 3 °C with a maximum computation time of 5.38 s. With the results obtained, the research effect provides a feasible method to control multiple input variables of an AHT, from a desired COP value. By properly supplying the waste heat to the AHT, it is possible to minimize the energy consumption and experimentation time, since the proposed method determines the best scenario to obtain good results, instead of performing different experimental tests. Finally the GA solved satisfactorily the ANNi multivariable function proposed, making feasible the use of this tool to optimize different variables at the same time.
In this research, global optimization algorithms were applied to solve the inverse artificial neural network (ANNi) for obtaining the best inputs values of an absorption heat transformer with energy ...recycling (AHTER) and improving its performance. The ANNi was obtained by inverting an artificial neural network (ANN) which architecture was 16 input variables, 3 neurons in the hidden layer and 1 output variable. The ANNi’s aim was optimizing 1, 2, 3, and up to 4 manipulated input variables, as well as calculating the other 12 input variables not manipulated in the system (AHTER) considering a coefficient of performance (COP) desired. The Cuckoo Search (CS), Particle Swarm Optimization (PSO), Genetic Algorithm (GA) and Simulated Annealing (SA) algorithms were used to find the optimal inputs. The results showed that the four algorithms used (ANNi-CS, ANNi-PSO, ANNi-GA, and ANNi-SA) satisfactorily optimize of 1 up to 16 inputs of the ANNi. However, the algorithms of ANNi-CS and ANNi-SA were slightly faster with acceptable accuracy. Additionally, they were carried out two analyses using different COPs values. These analyses showed that both algorithms optimize the AHTER’s inputs for different COP, as well as R>0.988 were obtained with the COP experimental data against COP obtained data by both ANNi models.
•4 global optimization algorithms were applied to solve the ANNi.•A multi-variable function was solved by CS, PSO, GA and SA.•1 to up 4 inputs variables were optimized to improve the COP of an AHTER.•1 to 16 inputs variables were obtained with ANNi-CS, ANNi-PSO, ANNi-GA, and ANNi-SA.•COP was maximized applying ANNi-CS and ANNi-SA considering the limits of the device.
An artificial neural network (ANN) model was developed to predict the coefficient of performance (COP) of an absorption heat transformer with a new physical design consisting of compact components, ...and its inverse (ANNi) was used to optimize the system's performance, coupled for the water purification. This ANN model takes into account the input and output temperature (T) of each duplex component (generator-condenser and evaporator-absorber), as well as the concentration of solution LiBr–H2O (X), pressure (P) and mass flow (ṁ). The best fitting training data was acquired with 16–7–1 considering a hyperbolic tangent sigmoid transfer-function in the hidden layer and a linear transfer-function in the output. Comparing the predicted and experimental data it was observed a satisfactory agreement (R2 > 0.9969 and MPE ∼ 3%). Furthermore, from this ANN model, a strategy was developed for optimization of a generator and an evaporator input temperatures using inverse artificial neural networks (ANNi) and solved by the method of genetic algorithms (GAs). The good prediction of the ANN model, as well as the optimized data using ANNi-GAs, makes it possible to control on-line the operation of the system, increasing the value of COP.
•Two duplex components were integrated as a new physical design for AHT.•ANN was developed to predict the COP performance.•Parameters were identified for optimal performance through inverse ANN.•Inverse ANN was solved with genetic algorithms.
The calculation of the performance of absorption heat transformers (AHTs) depends on multiple variables. In this work, artificial neural network (ANN) models with new configurations were developed to ...simultaneously estimate the coefficient of performance (COP) and Carnot coefficient of performance (COP
Carnot
) of an AHT prototype. The variables used to train the models were: the inlet and outlet temperatures corresponding to the main components of the AHT. The output parameters to simulate were the COP and COP
carnot
, which are important values to determine the performance and real efficiency based on the Carnot cycle, respectively. To find the appropriate model, it was necessary to explore learning algorithms, activation functions, and multilayers. The results show a good estimation of the output parameters through three configurations of the ANN model. However, based on the number of coefficients obtained during learning and the simultaneous simulation of two output parameters, a multilayer ANN model was proposed as the best configuration. Therefore, an architecture of four neurons in the first hidden layer and four neurons in the second hidden layer (08:04:04:02) was sufficient to reproduce the output parameters, achieving a value of
R
2
of 0.9265, 0.9573 and with a mean absolute percentage error of 2.41, 1.14% for COP and COP
Carnot
, respectively. In the three configurations, the use of hyperbolic tangent sigmoid activation function (TANSIG) in the hidden layers and the adjustment of the coefficients with the Levenberg–Marquardt learning algorithm obtained the best results. The influence of each of the variables selected for the ANN model was analyzed through a correlation matrix and a sensitivity analysis. Other experimental variables were added in the training of the ANN model to consult the impact caused during the simultaneous prediction of the performance coefficients.