Gust response alleviation is very important for helicopters which have strong coupling and vibration. Gust disturbance not only influences the ride quality and the precision of the weapon delivery, but also affects to...Gust response alleviation is very important for helicopters which have strong coupling and vibration. Gust disturbance not only influences the ride quality and the precision of the weapon delivery, but also affects to the structural fatigue load and the strength. The method of an optimal control law to suppress the gust disturbance for helicopters is presented. The optimization requires the minimization of the vertical overload at the pilot′s seat, the attitude variation and the control energy consumption under the gust disturbance. Based on the original control system, the new system can be easily realized by adding a vertical speed feedback passage. In order to develop the real-time operational flight control system, the optimized control law is written in C language. The hybrid simulations prove that the performance of gust response alleviation and the efficiency of digitalization are satisfactory.展开更多
Aim To put forward a type of math model for optimizing fan′s twisting law.Methods This math model wu based on turbo-machinery Euler equations and calculus of variation, it was conducted for optimizing the aerodynamic...Aim To put forward a type of math model for optimizing fan′s twisting law.Methods This math model wu based on turbo-machinery Euler equations and calculus of variation, it was conducted for optimizing the aerodynamic parameters along the blade height of the fan, and the math method was produced for the optimization of fan's twisting law. Results The type 6102Q engine cooling fan was optimized by use of this model, and the calculation on data were contrasted with those of iso-reaction coefficiency flow type and free vortex flow type. Some probleme existing in long blade can be solved by use of above method. Conclusion The design parameters needn't be determined artificially, so calculating results are more rational to a high degree than that from other mehods.展开更多
The aim of the study is to investigate the absorption laws of nitrogen, phosphorus, and potassium, and proper nitrogen application in Chuanxiangyou 9838 under no-tillage cultivation. Five nitrogen application treatmen...The aim of the study is to investigate the absorption laws of nitrogen, phosphorus, and potassium, and proper nitrogen application in Chuanxiangyou 9838 under no-tillage cultivation. Five nitrogen application treatments were designed to analyze the absorption laws of N, P and K, and to discuss the effects of different N fertilizer application amounts on yield and yield composition factors of Chuanxiangyou 9838. The results showed that gross nutrient absorption in Chuanxiangyou 9838 was greatly varied at different developmental stages under rice-rape rotation with no-tillage. The maximum N absorption in Chuanxiangyou 9838 appeared at jointing stage followed by heading stage, thirdly the tillefing stage ; the P absorption in Chuanxiangyou 9838 presented the consecutively slight increase during seedling stage and mature stage ; the K absorption in Chuanxiangyou 9838 was mainly conducted from jointing stage to heading stage, during which K absorption accounts for 73.4% of the total absorption in whole developmental stage. Consequently, N fertilizer should be applied earlier ( before jointing stage), P fertilizer is suitable as base fertilizer and application of K fertilizer should be preferably conducted at early-middle period. When the yield reached 11 t/hm^2, the optimal N application amount in Chuanxiangyou 9838 was about 165 kg/hm^2.展开更多
It is generally impossible to obtain the analytic optimal guidance law for complex nonlinear guidance systems of homing missiles,and the open loop optimal guidance law is often obtained by numerical methods,which can ...It is generally impossible to obtain the analytic optimal guidance law for complex nonlinear guidance systems of homing missiles,and the open loop optimal guidance law is often obtained by numerical methods,which can not be used directly in practice.The neural networks are trained off line using the optimal trajectory of the missile produced by the numerical open loop optimal guidance law,and then,the converged neural networks are used on line as the feedback optimal guidance law in real time.The research shows that different selections of the neural networks inputs,such as the system state variables or the rate of LOS(line of sight),may have great effect on the performances of the guidance systems for homing missiles.The robustness for several guidance laws is investigated by simulations,and the modular neural networks architectures are used to increase the approximating and generalizing abilities in the large state space.Some useful conclusions are obtained by simulation results.展开更多
To get better tracking performance of attitude command over the reentry phase of vehicles, the use of state-dependent Riccati equation (SDRE) method for attitude controller design of reentry vehicles was investigated....To get better tracking performance of attitude command over the reentry phase of vehicles, the use of state-dependent Riccati equation (SDRE) method for attitude controller design of reentry vehicles was investigated. Guidance commands are generated based on optimal guidance law. SDRE control method employs factorization of the nonlinear dynamics into a state vector and state dependent matrix valued function. State-dependent coefficients are derived based on reentry motion equations in pitch and yaw channels. Unlike constant weighting matrix Q, elements of Q are set as the functions of state error so as to get satisfactory feedback and eliminate state error rapidly, then formulation of SDRE is realized. Riccati equation is solved real-timely with Schur algorithm. State feedback control law u(x) is derived with linear quadratic regulator (LQR) method. Simulation results show that SDRE controller steadily tracks attitude command, and impact point error of reentry vehicle is acceptable. Compared with PID controller, tracking performance of attitude command using SDRE controller is better with smaller control surface deflection. The attitude tracking error with SDRE controller is within 5°, and the control deflection is within 30°.展开更多
A mathematical model of the soil pressure system in shield tunneling was proposed to optimize soil pressure control in the soil chamber, based on the constitutive relationship between strain and stress. The desired pr...A mathematical model of the soil pressure system in shield tunneling was proposed to optimize soil pressure control in the soil chamber, based on the constitutive relationship between strain and stress. The desired pressure is determined by using the finite element method. A linear quadratic constant state tracking problem was considered over an infinite time interval. The optimal control law was derived by differentiating the Hamilton function with respect to system input. In order to verify the effectiveness of the proposed mathematical model and optimal control law, an experimental study on the pressure control of the soil chamber in shield tunneling was conducted in a laboratory. The experiment results show that soil pressure in the soil chamber in shield tunneling can be accurately controlled.展开更多
The finite time thermodynamic performance of a generalized Carnot cycle, in which the heat transfer between the working fluid and the heat reservoirs obeys the generalized law Q∝( Δ T) m , is studied. The optimal ...The finite time thermodynamic performance of a generalized Carnot cycle, in which the heat transfer between the working fluid and the heat reservoirs obeys the generalized law Q∝( Δ T) m , is studied. The optimal configuration and the fundamental optimal relation between power and efficiency of the cycle are derived. Some special examples are discussed. The results can provide some theoretical guidance for the design a practical engine.展开更多
A dynamic programming-sequential quadratic programming(DP-SQP)combined algorithm is proposed to address the problem that the traditional continuous control method has high computational complexity and is easy to fall ...A dynamic programming-sequential quadratic programming(DP-SQP)combined algorithm is proposed to address the problem that the traditional continuous control method has high computational complexity and is easy to fall into local optimal solution.To solve the globally optimal control law sequence,we use the dynamic programming algorithm to discretize the separation control decision-making process into a series of sub-stages based on the time characteristics of the separation allocation model,and recursion from the end stage to the initial stage.The sequential quadratic programming algorithm is then used to solve the optimal return function and the optimal control law for each sub-stage.Comparative simulations of the combined algorithm and the traditional algorithm are designed to validate the superiority of the combined algorithm.Aircraft-following and cross-conflict simulation examples are created to demonstrate the combined algorithm’s adaptability to various conflict scenarios.The simulation results demonstrate the separation deploy strategy’s effectiveness,efficiency,and adaptability.展开更多
This paper aims to discuss the development and functioning conditions of business networks. After recalling the main characteristics of post-fordistic environment and comparing it to a "stormy sea" (section one) o...This paper aims to discuss the development and functioning conditions of business networks. After recalling the main characteristics of post-fordistic environment and comparing it to a "stormy sea" (section one) of the paper focuses on the idea of networks described as "rafts" useful to firms to build their own competitive advantages. In fact, while theoretical knowledge is not so valuable because everybody can have it, practical and contextual knowledge is specific and therefore it can be defended. The development of a contextual knowledge is feasible if the firm chooses among all the possible alternatives. Subsequently in section two, it shows how fordistic principles eliminate space, reduce time, and increase the speed of communication among individuals and as entering a network has become a necessity as it allows a firm to obtain competitive advantages. The greatest benefit is the chance to share the task of creating new knowledge among different members. In section three it is discussed if navigation in the post-fordistic stormy sea could take advantage from the existence of a more certain regulation. It is necessary to underline that positive law is not a post-fordistic tool. There is no satisfactory detailed law regarding ideas, knowledge, and know-how, by now. Therefore, it is not possible to rely on a specific regulation framework to protect knowledge found on the network. In conclusion in section four, the work discusses how single organizations need to reach the "raft"--which is the network--through the idea of sharing learning and distinguishing elements necessary to survive in the stormy sea post-Fordism environment. Lastly, section five would be analyzed a public institution--Milan Chamber of Commerce--which has "changed its dress" to more effectively perform its support role to firms.展开更多
Research has been conducted about the hardness prediction for the carburizing and quenching process based on an optimized hardness simulation model,in accordance with the calculation rule of mixed phases.The coupling ...Research has been conducted about the hardness prediction for the carburizing and quenching process based on an optimized hardness simulation model,in accordance with the calculation rule of mixed phases.The coupling field model incorporates carburizing field analysis,temperature field analysis,phase transformation kinetics analysis and a modified hardness calculation model.In determination of the calculation model for hardness,calculation equations are given to be applied to low carbon content(x(C)<0.5%) for the child phases and the martensite hardness is calculated for high carbon content(x(C)>0.5%) in alloy.Then,the complete carburizing-quenching hardness calculation model is built,and the hardness simulation data are corrected considering the influence of residual austenite(RA) on hardness.Hardness simulations of the carburizing and quenching process of 17CrNiMo6 samples have been performed using DEFORM-HT_V10.2 and MATLAB R2013 a.Finally,a series of comparisons of simulation results and measured values show a good agreement between them,which validates the accuracy of the proposed mathematical model.展开更多
This essay poses Walras's theory of price mechanism in its merits and limitations. Walras proposed two laws as conditions for general equilibrium, namely: (1) the law of the variation of equilibrium prices, a subj...This essay poses Walras's theory of price mechanism in its merits and limitations. Walras proposed two laws as conditions for general equilibrium, namely: (1) the law of the variation of equilibrium prices, a subjective condition; and (2) the law of the establishment of equilibrium prices, an objective condition. Walras jointed both laws in order to develop his law of supply and demand. This paper offers a formal Walrasian approximation in terms of the Lyapounov's function, taking the diagonal dominant hypothesis as departure point, rediscovered almost a century after it was originally proposed by Walras. The paper concludes with critical reflection concerning the idea of equilibrium economics as medium of social cohesion.展开更多
Shanghai went into the ranks of the aging society in 1979, as the first area which entered into the aging society in China. Along with the arrival of the ageing, the nursing problems of the old man and disabled elderl...Shanghai went into the ranks of the aging society in 1979, as the first area which entered into the aging society in China. Along with the arrival of the ageing, the nursing problems of the old man and disabled elderly become the important factors which affect social development. The establishment of the legal system, System integration to realize resource optimal allocation, Division of multilevel optimization services provide new pattern can make it happen.展开更多
This paper investigates the MED (Minimum Entransy Dissipation) optimization of heat transfer processes with the generalized heat transfer law q ∝ (A(T^n))m. For the fixed amount of heat transfer, the optimal te...This paper investigates the MED (Minimum Entransy Dissipation) optimization of heat transfer processes with the generalized heat transfer law q ∝ (A(T^n))m. For the fixed amount of heat transfer, the optimal temperature paths for the MED are obtained The results show that the strategy of the MED with generalized convective law q ∝ (△T)^m is that the temperature difference keeps constant, which is in accordance with the famous temperature-difference-field uniformity principle, while the strategy of the MED with linear phenomenological law q ∝ A(T^-1) is that the temperature ratio keeps constant. For special cases with Dulong-Petit law q ∝ (△T)^1.25 and an imaginary complex law q ∝ (△(T^4))^1.25, numerical examples are provided and further compared with the strategies of the MEG (Minimum Entropy Generation), CHF (Constant Heat Flux) and CRT (Constant Reservoir Temperature) operations. Besides, influences of the change of the heat transfer amount on the optimization results with various heat resistance models are discussed in detail.展开更多
In this paper, a novel iterative Q-learning algorithm, called "policy iteration based deterministic Qlearning algorithm", is developed to solve the optimal control problems for discrete-time deterministic no...In this paper, a novel iterative Q-learning algorithm, called "policy iteration based deterministic Qlearning algorithm", is developed to solve the optimal control problems for discrete-time deterministic nonlinear systems. The idea is to use an iterative adaptive dynamic programming(ADP) technique to construct the iterative control law which optimizes the iterative Q function. When the optimal Q function is obtained, the optimal control law can be achieved by directly minimizing the optimal Q function, where the mathematical model of the system is not necessary. Convergence property is analyzed to show that the iterative Q function is monotonically non-increasing and converges to the solution of the optimality equation. It is also proven that any of the iterative control laws is a stable control law. Neural networks are employed to implement the policy iteration based deterministic Q-learning algorithm, by approximating the iterative Q function and the iterative control law, respectively. Finally, two simulation examples are presented to illustrate the performance of the developed algorithm.展开更多
Examples of heat transfer and heat-work conversion are optimized with entropy generation and entransy loss,respectively based on the generalized heat transfer law in this paper.The applicability of entropy generation ...Examples of heat transfer and heat-work conversion are optimized with entropy generation and entransy loss,respectively based on the generalized heat transfer law in this paper.The applicability of entropy generation and entransy loss evaluation in these optimization problems is analyzed and discussed.The results show that the entransy loss rate reduces to the entransy dissipation rate in heat transfer processes,and that the entransy loss evaluation is effective for heat transfer optimization.However,the maximum heat transfer rate does not correspond to the minimum entropy generation rate with prescribed heat transfer temperature difference,which indicates that the entropy generation minimization is not always appropriate to heat transfer optimization.For heat-work conversion processes,the maximum entransy loss rate and the minimum entropy generation rate both correspond to the maximum output power,and they are both appropriate to the optimization of the heat-work conversion processes discussed in this paper.展开更多
文摘Gust response alleviation is very important for helicopters which have strong coupling and vibration. Gust disturbance not only influences the ride quality and the precision of the weapon delivery, but also affects to the structural fatigue load and the strength. The method of an optimal control law to suppress the gust disturbance for helicopters is presented. The optimization requires the minimization of the vertical overload at the pilot′s seat, the attitude variation and the control energy consumption under the gust disturbance. Based on the original control system, the new system can be easily realized by adding a vertical speed feedback passage. In order to develop the real-time operational flight control system, the optimized control law is written in C language. The hybrid simulations prove that the performance of gust response alleviation and the efficiency of digitalization are satisfactory.
文摘Aim To put forward a type of math model for optimizing fan′s twisting law.Methods This math model wu based on turbo-machinery Euler equations and calculus of variation, it was conducted for optimizing the aerodynamic parameters along the blade height of the fan, and the math method was produced for the optimization of fan's twisting law. Results The type 6102Q engine cooling fan was optimized by use of this model, and the calculation on data were contrasted with those of iso-reaction coefficiency flow type and free vortex flow type. Some probleme existing in long blade can be solved by use of above method. Conclusion The design parameters needn't be determined artificially, so calculating results are more rational to a high degree than that from other mehods.
基金Project of Scientific and Technical Supporting Programs (2006AD05B06)Key Technologies Research and Development Program of Sichuan Province during the 10th Five-year Plan During the 11th Five-year Plan(2006YZGG-28)the project from International Plant Nutrition Institute (IPNI)~~
文摘The aim of the study is to investigate the absorption laws of nitrogen, phosphorus, and potassium, and proper nitrogen application in Chuanxiangyou 9838 under no-tillage cultivation. Five nitrogen application treatments were designed to analyze the absorption laws of N, P and K, and to discuss the effects of different N fertilizer application amounts on yield and yield composition factors of Chuanxiangyou 9838. The results showed that gross nutrient absorption in Chuanxiangyou 9838 was greatly varied at different developmental stages under rice-rape rotation with no-tillage. The maximum N absorption in Chuanxiangyou 9838 appeared at jointing stage followed by heading stage, thirdly the tillefing stage ; the P absorption in Chuanxiangyou 9838 presented the consecutively slight increase during seedling stage and mature stage ; the K absorption in Chuanxiangyou 9838 was mainly conducted from jointing stage to heading stage, during which K absorption accounts for 73.4% of the total absorption in whole developmental stage. Consequently, N fertilizer should be applied earlier ( before jointing stage), P fertilizer is suitable as base fertilizer and application of K fertilizer should be preferably conducted at early-middle period. When the yield reached 11 t/hm^2, the optimal N application amount in Chuanxiangyou 9838 was about 165 kg/hm^2.
文摘It is generally impossible to obtain the analytic optimal guidance law for complex nonlinear guidance systems of homing missiles,and the open loop optimal guidance law is often obtained by numerical methods,which can not be used directly in practice.The neural networks are trained off line using the optimal trajectory of the missile produced by the numerical open loop optimal guidance law,and then,the converged neural networks are used on line as the feedback optimal guidance law in real time.The research shows that different selections of the neural networks inputs,such as the system state variables or the rate of LOS(line of sight),may have great effect on the performances of the guidance systems for homing missiles.The robustness for several guidance laws is investigated by simulations,and the modular neural networks architectures are used to increase the approximating and generalizing abilities in the large state space.Some useful conclusions are obtained by simulation results.
基金Project(51105287)supported by the National Natural Science Foundation of China
文摘To get better tracking performance of attitude command over the reentry phase of vehicles, the use of state-dependent Riccati equation (SDRE) method for attitude controller design of reentry vehicles was investigated. Guidance commands are generated based on optimal guidance law. SDRE control method employs factorization of the nonlinear dynamics into a state vector and state dependent matrix valued function. State-dependent coefficients are derived based on reentry motion equations in pitch and yaw channels. Unlike constant weighting matrix Q, elements of Q are set as the functions of state error so as to get satisfactory feedback and eliminate state error rapidly, then formulation of SDRE is realized. Riccati equation is solved real-timely with Schur algorithm. State feedback control law u(x) is derived with linear quadratic regulator (LQR) method. Simulation results show that SDRE controller steadily tracks attitude command, and impact point error of reentry vehicle is acceptable. Compared with PID controller, tracking performance of attitude command using SDRE controller is better with smaller control surface deflection. The attitude tracking error with SDRE controller is within 5°, and the control deflection is within 30°.
基金Supported by the National Basic Research Project (2007CB714006, 90815023) the National Natural Science Foundation of China (GZ0818, GZ1107)
文摘A mathematical model of the soil pressure system in shield tunneling was proposed to optimize soil pressure control in the soil chamber, based on the constitutive relationship between strain and stress. The desired pressure is determined by using the finite element method. A linear quadratic constant state tracking problem was considered over an infinite time interval. The optimal control law was derived by differentiating the Hamilton function with respect to system input. In order to verify the effectiveness of the proposed mathematical model and optimal control law, an experimental study on the pressure control of the soil chamber in shield tunneling was conducted in a laboratory. The experiment results show that soil pressure in the soil chamber in shield tunneling can be accurately controlled.
文摘The finite time thermodynamic performance of a generalized Carnot cycle, in which the heat transfer between the working fluid and the heat reservoirs obeys the generalized law Q∝( Δ T) m , is studied. The optimal configuration and the fundamental optimal relation between power and efficiency of the cycle are derived. Some special examples are discussed. The results can provide some theoretical guidance for the design a practical engine.
基金supported in part by the National Natural Science Foundation of China(Nos.61773202,52072174)the Foundation of National Defense Science and Technology Key Laboratory of Avionics System Integrated Technology of China Institute of Aeronautical Radio Electronics(No.6142505180407)+1 种基金the Open Fund for Civil Aviation General Aviation Operation Key Laboratory of China Civil Aviation Management Cadre Institute(No.CAMICKFJJ-2019-04)the National key R&D plan(No.2021YFB1600500)。
文摘A dynamic programming-sequential quadratic programming(DP-SQP)combined algorithm is proposed to address the problem that the traditional continuous control method has high computational complexity and is easy to fall into local optimal solution.To solve the globally optimal control law sequence,we use the dynamic programming algorithm to discretize the separation control decision-making process into a series of sub-stages based on the time characteristics of the separation allocation model,and recursion from the end stage to the initial stage.The sequential quadratic programming algorithm is then used to solve the optimal return function and the optimal control law for each sub-stage.Comparative simulations of the combined algorithm and the traditional algorithm are designed to validate the superiority of the combined algorithm.Aircraft-following and cross-conflict simulation examples are created to demonstrate the combined algorithm’s adaptability to various conflict scenarios.The simulation results demonstrate the separation deploy strategy’s effectiveness,efficiency,and adaptability.
文摘This paper aims to discuss the development and functioning conditions of business networks. After recalling the main characteristics of post-fordistic environment and comparing it to a "stormy sea" (section one) of the paper focuses on the idea of networks described as "rafts" useful to firms to build their own competitive advantages. In fact, while theoretical knowledge is not so valuable because everybody can have it, practical and contextual knowledge is specific and therefore it can be defended. The development of a contextual knowledge is feasible if the firm chooses among all the possible alternatives. Subsequently in section two, it shows how fordistic principles eliminate space, reduce time, and increase the speed of communication among individuals and as entering a network has become a necessity as it allows a firm to obtain competitive advantages. The greatest benefit is the chance to share the task of creating new knowledge among different members. In section three it is discussed if navigation in the post-fordistic stormy sea could take advantage from the existence of a more certain regulation. It is necessary to underline that positive law is not a post-fordistic tool. There is no satisfactory detailed law regarding ideas, knowledge, and know-how, by now. Therefore, it is not possible to rely on a specific regulation framework to protect knowledge found on the network. In conclusion in section four, the work discusses how single organizations need to reach the "raft"--which is the network--through the idea of sharing learning and distinguishing elements necessary to survive in the stormy sea post-Fordism environment. Lastly, section five would be analyzed a public institution--Milan Chamber of Commerce--which has "changed its dress" to more effectively perform its support role to firms.
基金Projects(51535012,U1604255)supported by the National Natural Science Foundation of ChinaProject(2016JC2001)supported by the Key Research and Development Program of Hunan Province,China
文摘Research has been conducted about the hardness prediction for the carburizing and quenching process based on an optimized hardness simulation model,in accordance with the calculation rule of mixed phases.The coupling field model incorporates carburizing field analysis,temperature field analysis,phase transformation kinetics analysis and a modified hardness calculation model.In determination of the calculation model for hardness,calculation equations are given to be applied to low carbon content(x(C)<0.5%) for the child phases and the martensite hardness is calculated for high carbon content(x(C)>0.5%) in alloy.Then,the complete carburizing-quenching hardness calculation model is built,and the hardness simulation data are corrected considering the influence of residual austenite(RA) on hardness.Hardness simulations of the carburizing and quenching process of 17CrNiMo6 samples have been performed using DEFORM-HT_V10.2 and MATLAB R2013 a.Finally,a series of comparisons of simulation results and measured values show a good agreement between them,which validates the accuracy of the proposed mathematical model.
文摘This essay poses Walras's theory of price mechanism in its merits and limitations. Walras proposed two laws as conditions for general equilibrium, namely: (1) the law of the variation of equilibrium prices, a subjective condition; and (2) the law of the establishment of equilibrium prices, an objective condition. Walras jointed both laws in order to develop his law of supply and demand. This paper offers a formal Walrasian approximation in terms of the Lyapounov's function, taking the diagonal dominant hypothesis as departure point, rediscovered almost a century after it was originally proposed by Walras. The paper concludes with critical reflection concerning the idea of equilibrium economics as medium of social cohesion.
文摘Shanghai went into the ranks of the aging society in 1979, as the first area which entered into the aging society in China. Along with the arrival of the ageing, the nursing problems of the old man and disabled elderly become the important factors which affect social development. The establishment of the legal system, System integration to realize resource optimal allocation, Division of multilevel optimization services provide new pattern can make it happen.
基金supported by the National Natural Science Foundation of China(Grant Nos.51576207,51356001&51579244)
文摘This paper investigates the MED (Minimum Entransy Dissipation) optimization of heat transfer processes with the generalized heat transfer law q ∝ (A(T^n))m. For the fixed amount of heat transfer, the optimal temperature paths for the MED are obtained The results show that the strategy of the MED with generalized convective law q ∝ (△T)^m is that the temperature difference keeps constant, which is in accordance with the famous temperature-difference-field uniformity principle, while the strategy of the MED with linear phenomenological law q ∝ A(T^-1) is that the temperature ratio keeps constant. For special cases with Dulong-Petit law q ∝ (△T)^1.25 and an imaginary complex law q ∝ (△(T^4))^1.25, numerical examples are provided and further compared with the strategies of the MEG (Minimum Entropy Generation), CHF (Constant Heat Flux) and CRT (Constant Reservoir Temperature) operations. Besides, influences of the change of the heat transfer amount on the optimization results with various heat resistance models are discussed in detail.
基金supported in part by National Natural Science Foundation of China(Grant Nos.6137410561233001+1 种基金61273140)in part by Beijing Natural Science Foundation(Grant No.4132078)
文摘In this paper, a novel iterative Q-learning algorithm, called "policy iteration based deterministic Qlearning algorithm", is developed to solve the optimal control problems for discrete-time deterministic nonlinear systems. The idea is to use an iterative adaptive dynamic programming(ADP) technique to construct the iterative control law which optimizes the iterative Q function. When the optimal Q function is obtained, the optimal control law can be achieved by directly minimizing the optimal Q function, where the mathematical model of the system is not necessary. Convergence property is analyzed to show that the iterative Q function is monotonically non-increasing and converges to the solution of the optimality equation. It is also proven that any of the iterative control laws is a stable control law. Neural networks are employed to implement the policy iteration based deterministic Q-learning algorithm, by approximating the iterative Q function and the iterative control law, respectively. Finally, two simulation examples are presented to illustrate the performance of the developed algorithm.
基金supported by the Natural Science Foundation of China(Grant No. 51136001)the Tsinghua University Initiative ScientificResearch Program
文摘Examples of heat transfer and heat-work conversion are optimized with entropy generation and entransy loss,respectively based on the generalized heat transfer law in this paper.The applicability of entropy generation and entransy loss evaluation in these optimization problems is analyzed and discussed.The results show that the entransy loss rate reduces to the entransy dissipation rate in heat transfer processes,and that the entransy loss evaluation is effective for heat transfer optimization.However,the maximum heat transfer rate does not correspond to the minimum entropy generation rate with prescribed heat transfer temperature difference,which indicates that the entropy generation minimization is not always appropriate to heat transfer optimization.For heat-work conversion processes,the maximum entransy loss rate and the minimum entropy generation rate both correspond to the maximum output power,and they are both appropriate to the optimization of the heat-work conversion processes discussed in this paper.