This paper describes design and verification of a wireless-interface SoC (system-on-a-chip) for a wireless battery-less mouse with short-range data-communication capability. The SoC comprises an RF transmitter and microcontroller. The SoC, which is powered by an electric generator that exploits gyration energy by dragging the mouse, was fabricated using a TSMC 0.18-um CMOS process. The features of the SoC are the adoption of a simple FSK modulation scheme, single-end configuration on the RF transmitter, and specific microcontroller design for mouse operation. We verified that the RF transmitter can make data communication within a 1-m range at 2.17 mW, and the microcontroller consumes 0.03 mW at 1 MHz, which exhibits that the total power consumption is 2.2 mW. This is sufficiently low for the SoC to operate with energy harvesting.
This paper describes an assimilation phenomenon on monochrome images divided from the size of patterns, and proposes an analytical model about this phenomenon. The frequency spectrum of a sample pattern by the discrete Fourier Transform and human visual characteristics by the Butterworth type Low Pass Filter (LPF) are formulized. A two-dimensional model of the relation between the samples and human visual characteristics, which is additionally introduced the concept of a noise level is proposed, and discussed with the assimilation phenomenon qualitatively and quantitatively. Then, a comparative examination of this model is carried out by subjective evaluation, using the visual distance, and the validity is verified. As a result, the analysis using this proposal model, the cut-off frequency of this LPF is set to 20 cycle/degree, its attenuance is set to -20 dB/decade, and the noise level presumes -50 dB. The results almost conformed qualitatively to the results of evaluation experiments in this example.
In this study, we address to quantify the relationship between the significant arm-clutch loading, leg restriction and motor paralysis, and analyze lumbar joint trajectories in the orthotic gait of paraplegic subjects and the ordinary and orthotic gaits of a normal subject using an inverted pendulum model. For the leg restriction, the trajectories are located in front of an equilibrium point of the inverted pendulum, and the loading is higher due to the influence of gravity moment. Comparing the trajectory of paraplegic and normal gait with orthosis in horizontal plane, the trajectory in the paraplegic subjects was rectilinear shape, while that in normal subject was curved in the direction to the equilibrium point. The loading is lower in the curved trajectory than in the straight trajectory because of the trade-off between gravity and inertia. These results suggest that the increase of the distance between the trunk movement and the equilibrium point of the inverted pendulum result in the significant loading due to the leg restriction and motor paralysis in orthotic gait of paraplegics.
We proposed the automatic license plate recognition system which based on artificial neural networks(ANNs). The targets images of the system were taken with fixed camera. In this paper, we applied the ANNs method for the images that were taken with onboard camera. The first evaluation is that the system recognizes positions of license plates that are shown in front of a car. We introduce a DP matching method to improve the segmentation of characters.
The conventional methods to detect the DTMF signals are accomplished by comparing the power of an input signal using the DFT. This paper proposes a new DTMF signals detector using a multiplier-less resonators which have not the multiplications to get output signals. The proposed method detects the DTMF signals using whether the outputs of the multiplier-less resonators resonate or not. This processing is realized by controlling a downsampling ratio, that is, a time component of the input signal. Therefore, the simple algorithm which can mainly be processed by the additions and subtractions is derived. We compare the proposed method for the DTMF signals detector with the conventional method in terms of the computational complexity and show the detection performance of the proposed method.
In this paper, we propose a new controller reduction method using the closed-loop block balanced realization technique. To decrease the closed-loop performance degradation, it is need to take account the closed-loop configuration in controller reduction process. The block balanced realization technique was developed from the open-loop balanced realization method to consider the closed-loop configuaration. In our method, we employ the block balanced realization method to reduce the order of H∞ controllers, and to consider robustness for controller variation, we use the general representation of H∞ controllers. A numerical example is presented to illustrate the effectiveness of our method.
Recently, a frequency selective adaptive filter design method for active noise control has been proposed based on the so-called least mean square algorithm. However, this method does not sufficiently exploit the degree of freedom of the step parameter of the recursive rule in the case where a priori information on an uncertain plant is available. This paper proposes a design method of the step parameter such that noise cancellation is guaranteed against the plant uncertainty. The design problem of the step parameter is reduced to an optimization problem involving linear matrix inequalities and is efficiently solvable. Experimental results are provided to illustrate the effectiveness of the proposed method.
This paper addresses a swing-up and stabilization problem of a cart-pendulum system using energy control and controlled Lagrangian methods. A common strategy for this problem is to switch two controllers for swing-up and stabilization according to the angle of the pendulum. In most cases, control performance greatly depends on the switching angle, and choice of a bad angle cannot even stand the pendulum upward. Thus, a lot of trial and error is necessary to find effective switching angles. To release us from the difficulty in choosing the switching angle, this paper proposes a robust control method to assist our choice based on controlled Lagrangian and energy control methods. Experimental results demonstrate that the proposed combination methods can effectively swing up and stabilize the pendulum, whereas existing methods cannot.
When a neural network simulates a Turing machine, the states of finite state controller and the symbols on infinite tape are encoded in continuous numbers of neuron's outputs. The precision of outputs is regarded as a space resource in neural computations. We show a sufficient condition about the precision to guarantee the correctness of computations. Linear precision suffice in regard to nT, where n is the number of neurons and T is the iteration count of state updates.
The target of this research is a rectangular cutting stock problem against flawed stock rolls, that orders are assigned to the stock rolls avoiding flaws and connecting the stock rolls for efficient assignment. The problem has the feature that total area of all stock rolls is larger than one of orders, so that, first the stock rolls used for assignment are selected from all stock rolls. In order to effectively use the “cumbersome stock rolls", whose assignment evaluation tends to be small, in this research, two methods are proposed; a method of formulizing degree of a cumbersome stock roll, and a method to connect divided stock rolls effectively. In the former method, the degree is expressed from two factors, namely rate of flaws' area and complexity of flaws, and the parameter is added to a total evaluation. In the latter, “improvement degree" of a pair of divided stock rolls is introduced, and the pair whose improvement degree is high, is connected. The proposed method has been applied to real cutting stock problems. It is confirmed that the method can generate the solutions as same level as expert's solutions.
A community currency is local money that is issued by local governments or Non-Profit Organization (NPO) to support social services. The purpose of introducing community currencies is to regenerate communities by fostering mutual aids among community members. In this paper, we propose a community currency trading method through partial intermediary process, under operational environments without introducing coordinators all the time. In this method, coordinators perform coordination between service users and service providers during several months from the start point of transactions. After the period of coordination, participants spontaneously make transactions based on their trust area and a trust evaluation method based on the number of provided services and complaint information. This method is especially effective to communities with close social networks and low trustworthiness. The proposed method is evaluated through multi-agent simulation.
Noun compounds are frequently encountered construction in nature language processing (NLP), consisting of a sequence of two or more nouns which functions syntactically as one noun. The translation of noun compounds has become a major issue in Machine Translation (MT) due to their frequency of occurrence and high productivity. In our previous studies on Super-Function Based Machine Translation (SFBMT), we have found that noun compounds are very frequently used and difficult to be translated correctly, the overgeneration of noun compounds can be dangerous as it may introduce ambiguity in the translation. In this paper, we discuss the challenges in handling Japanese noun compounds in an SFBMT system, we present a shallow method for translating noun compounds by using a word level translation dictionary and target language monolingual corpus.
Real-coded genetic algorithms (GAs) are effective methods for function optimization. Generally speaking, the major crossover methods used in real-coded GAs require a large execution time for calculating the fitness of many children at each crossover. Thus, a new crossover method is needed for searching such a large search space efficiently. A novel crossover method that generates children stepwise is proposed and applied to the conventional generation-alternation model. In experiments based on standard test functions and actual problems, the proposed model found an optimal solution 30-40% faster than did the conventional model.
This paper presents a recommender system for TPO (Time, Place, and Occasion)-dependent goods. The TPO-dependent goods have three features: many attributes, multiformity, and high-frequency update. In order to recommend alternatives of the goods, our system (a) abstracts and metrizes the user's preference implied in sales records, and (b) filters massive alternatives by three kinds of methods: High-Angle Search, Low-Angle Search and Neighbor Search. Additionally, this paper describes the improvement method of the recommendation accuracy by memory-based reasoning with user's preference to latter two kinds of search. The numerical simulation for 10,000 user's data and 400,000 sales records has shown their accuracy.
The sterilization in the air was examined by passing the fungus between floating multi-electrode. This phenomenon seems to be because the fungus was destroyed in the strong electric field between electrodes. In this case, the oxidation sterilization does not occur, because the ozone is not generated.
A general technique is introduced to enhance the performance of carrier phase error synchronization loops. This technique employs a notch filter inserted in a structure cascaded with the loop filter. Computer simulations of some well-known carrier phase synchronization loops enhanced with the proposed technique show that an approximately jitter free tracking and a great improvement in bit error rate (BER) results have been achieved.