Search

Smart city construction is closely related to building energy consumption data analysis, and energy consumption prediction is helpful to guide urban power dispatching strategy. In general, the artificial neural network models used for energy consumption prediction are often based on the optimization of a single network, which has some shortcomings such as poor generalization ability and unstable fitting accuracy. Therefore, in order to improve the performance of BP neural network in building air conditioning energy consumption prediction, a per-column optimization prediction model is constructed in this paper. In order to avoid falling into the local minimum, genetic algorithm and particle swarm optimization algorithm are introduced to optimize the weight and threshold of the basic BP neural network to avoid the randomness of the parameters. Then, in order to improve the reliability and prediction accuracy of the model, the optimal prediction value is identified on the basis of the BP model prediction value based on the optimization algorithm. In addition, this paper makes an experimental study on the real air-conditioning energy consumption of a building in Laixi City, and the experimental results show the superiority of the model. The prediction accuracy is improved by about 79%, and the reliability of the algorithm is also improved. In the long run, the model can provide advance prediction for urban power dispatching and contribute to the construction of smart cities.

Dynamic thermal comfort control can provide comfortable, healthy, and energy-saving indoor environments. The hot / cold complaints occur when the indoor occupants are dissatisfied with the thermal environment. In order to obtain a precise index for thermal comfort control according to individual preferences, A PMV (Predicted Mean Vote)-based cold / hot complaint event method are presented so that the optimization control policies that affect both energy consumption and comfort can be formulated. A mathematical model of the number of cold and hot complaints in a time period t is developed that is based on the level-crossing theory of stochastic processes. Hot complaint experiments were conducted in intelligent environment laboratory. Experiment results demonstrate the effectiveness of the proposed method.

Comfortable, healthy, and energy-saving indoor environments can be obtained via a dynamic thermal comfort control. Difficulties to design an optimal control system for a dynamic thermal environment arise due to the lack of coordinative control evaluation methods for conflicting comfort and energy-saving indices. An improved multi-objective algorithm based on discrete PSO (Particle Swarm Optimization) is proposed to calculate the optimal values of parameters in the dynamic comfort control system based on users balance between the comfort and energy conservation. No a priori information or physical indoor environment model is needed. Experiment results demonstrate the effectiveness of the proposed control method.

In order to realize the data acquisition of field instruments, device control and data communication for GSHP (ground source heat pump) systems, a multi-functional CAN (Controller Area Network) bus controller is designed with a single-chip STC12C5A60S2 microcomputer as the core. The controller makes CAN communication, wireless communication and communication conversion from RS-485 to CAN possible. Moreover, data acquisition and control modules are designed, which include an A/D (Analog/Digital) converter module, a D/A (Digital/Analog) converter module and a DI/DO (Digital Input/Digital Output) module. Testing of the controller indicates that it has strong adaptability and flexibility. Therefore, this controller meets the different data acquisition and control needs for GSHP systems.

This paper is to realize the dynamic comfort control of indoor thermal comfort system using the semi-tensor product(STP) of matrices. A real-time fuzzy matrix can be obtained by analyzing the data, which can convert logical reasoning into digital signal processing. Then this air-conditioning fuzzy controller is applied to an indoor thermal comfort control system. Experiments and simulation results demonstrate that the air-conditioning fuzzy controller can be used to control indoor thermal comfort system by using the real-time fuzzy relation matrix.

A cerebellar model articulation controller (CMAC) is often used as nonlinear function approximator in pattern recognition, system identification, and neurocontrol systems. Generalization ability and learning accuracy are among important performance of a CMAC besides learning speed, and they are affected significantly by quantization level and the ratio of adjusted weights at each learning with a sample to all weights, called weigh adjusting ratio (WAR) in this paper. There is always a trade-off between generalizatinn ability and learning accuracy in conventional neural networks. In many cases, quantization level is increased to enhance learning accuracy without consideration of changing the WAR, this may result in decreasing generalization ability. In this paper, we studied Hyperball CMAC (HCMAC), and found the relation of generalization ability or learning accuracy with quantization level and WAR, a large number simulations illustrate that the CMAC is very effective in approximating nonlinear functions. On one hand, for fixed quantization level, the generalization error and learning error will conformably decrease with increasing WAR until WAR reaches a certain value, and they remain unchanged when the WAR is greater than the value. On the other hand, when the WAR fixed. learning error is decreasing with large quantization level, whereas generalization error has no clear relation with quantization level. These unique properties can be applied to improve the performance of designed CMAC. For further enhancing generalization ability and learning accuracy, a multiple CMAC structure (called parallel CMAC structure in this paper) is proposed based on the CMAC to approximate a highly nonlinear function very accurately. A simulation example demonstrates that the structure has powerful generalization and learning ability

The cerebellar model articulation controller (CMAC) is often used in neural computation and neurocontrol systems as output predictors or models of dynamic systems through static mapping from input space to output space. Hyperball CMAC (HCMAC) is presented in the paper. As a fast learning network, its performance is very easy to be analyzed. The simulation results, from the application of three kinds of CMAC, demonstrate that the HCMAC has a better learning performance than CMAC with general basis functions (Chiang; Lin, 1996) and general Albus CMAC (Albus, 1975). For mapping with large-dimensional input, the conventional CMACs need large weight memory space, while the need usually is not met. In order to overcome the issue, a new CMAC structure containing several HCMACs is proposed, the input space of each HCMAC is a subset of input space to be mapped. The CMAC structure is called SHCMAC. Simulations for the SHCMAC used in a multivariable nonlinear dynamic system modelling are performed to demonstrate its powerful associative memory performance and the accuracy in modeling

Using semi-tensor product (STP) of matrices, with observed datasets, this paper proposes a new and more general framework to construct a matrix-based fuzzy relation structure model for multi-input multi-output (MIMO) fuzzy control systems. The measured sampling data of inputs and outputs are assumed to be obtained from experiments. Instead of building the fuzzy logical rule sets of a fuzzy dynamical process, the algebraic form is constructed directly. The whole designing process can be realized via matrix expression and algebraic computing. Then this novel fuzzy controller is applied into a thermal comfort control system and works well with good controlled performance. The new technique proposes one general design method to obtain a fuzzy relation matrix expression of multiple variables fuzzy control systems. It is particularly suitable to design fuzzy controllers of non-decomposable multi-output systems, which is not solvable directly by the traditional decomposed control design methods.

The inhabited environments are MIMO, uncertainly, and nonlinear complex systems. This paper presents a novel intelligent fuzzy agent(IFA) based on input-output associational algorithm for intelligent inhabited environments. An input-output dynamic associational algorithm based on Hebb learning is proposed, which can divide a complex system into multiple simple systems and eliminate the irrelevant data from the learning data to improve the learning rate. The IFA learns the user's preferences according to the manually operation of the user and proactively controls the environments. Initially the IFA extracts the membership functions and fuzzy rules from the data captured. After that the IFA learns the membership functions and optimizes the fuzzy rules online when the user's preferences change. The experience results show that the proposed system is effective.

An intelligent inhabited environment applying interconnected embedded agents by network has intelligent reasoning, planning learning, and control capabilities. Thermal and light comforts are two major control objectives for the environment to deal with using data-driven control method. Practically, dynamic association level of agents should be learned from online data with three reasons: changing structure of agents with the devices to be added to or removed from the environment during residents' life, a large number of dimension of input and output vectors making it is very difficult to design learning based controller, and a multitude of interconnected embedded agents resulting in major load in network communication and calculation. This paper presented a novel online learning algorithm to obtain the structure agents with different functions through identifying the associations between inputs and outputs of the environment. An association weight matrix can be calculated online and the embedded agents can be dynamically divided into multiple subgroups. This can reduce dimension of input vector for each subgroup, reducing network communication load among embedded agents, decreasing the complexities of programming, and improving the learning rate of agents. The experiment results demonstrated the effectiveness and significance of the learning algorithm.

Dynamic thermal comfort control can effectively improve the levels of comfortable feeling, energy-saving, and human health. Traditional control theories and methods based on accurate physical model can not solve the control problems met during analyzing and controlling thermal comfort because of difficulties in obtaining the accurate model of a thermal environment. This paper presented a novel control strategy of dynamic thermal comfort using online and offline dada collected in thermal environment. Family members' fuzzy sensations are involved in the closed control loop. A six-input one-output CMAC neural network was used to learn the relationship between thermal variables and parameters and Predicted Mean Vote (PMV). This CMAC functions as a predictive PMV model. Another three-input one-output CMAC learned the nonlinear relationship between temperature decrease and humidity deduction in summer, as humidity deduction model when temperature changes in value. We also defined the concepts of personal and dynamic thermal comfort zones, proposing a fuzzy information fusion method of thermal sensation using fuzzy set theory. Procedure of computational experiments with these data-based predictive models above was designed to determine appropriate set points of thermal variables in order for thermal comfort value to periodically be within the desired personal comfort zone or dynamic thermal comfort zone. The control strategy given in the paper can help researchers design dynamic thermal comfort system in detail.

In intelligent inhabited environments(IIEs), the different user’s preferences and needs are different, and the same user’s preferences will change over time, which desire that the intelligent agents have evolvement function. In this paper, we propose a new fuzzy control system that is embedded in the intelligent agent which is a function agent in Multi-Agent System(MAS) based on ZigBee wireless sensor network. The system learns the user’s preferences according to the manually operation of the user and proactively controls the environment. The intelligent fuzzy agent includes four phases: (1) Capturing input-output data pairs. (2) Extracting fuzzy rules. (3) Agent control. (4). Modifying fuzzy rules online. Initially the system’s fuzzy rules are extracted from the collected information of sensors and actuators in IIEs, after that the system rapidly optimizes the fuzzy rules when the user’s preferences change. The experience results show that the proposed system is effective. Moreover, adding the time and user behaviors inputs will improve the learning accuracy of the system.

Designing less intrusive intelligent environments requires a deep understanding of activities that a user is engaged in. This paper presents a novel one-pass neural network system that uses unobtrusive and relatively simple sensors and puts forward a constructive algorithm which is able to recognize different high level activities (such as ldquosleepingrdquo, ldquowashingrdquo, ldquoworking at computerrdquo) in intelligent inhabited environments. The neural network system adding temporal capabilities is able to recognize abnormal behaviors. One-pass learning method of weight ratios can rapidly improve the learning speed and reduce the memory of embedded computer. It can be trained in an online mode and hence it can be integrated into the limited processor-power embedded computing platforms used in intelligent environments. Experiment results show that this method is transparent, simple and effective.