1. Introduction In recent years, the tourism industry is important for revitalizing local cities and improving the quality of life. In Tourism-Based Community, it is required to investigate tourism resources that will be attractive for nonresidents. Meanwhile, globalization is progressing now, and foreign tourists visiting Japan are also increasing. Therefore, in order to discover attractive tourism resources for nonresidents, it is necessary to investigate the needs of foreign tourists as nonresidents for Japanese tourist spots. In addition, tourism behavior is generally carried out in the process of actually visiting after obtaining prior information to determine the destination. Therefore, it’s necessary to investigate changes in the impression of tourists to tourist spots before and after visiting. In this study, to get a thorough understanding of attractive tourism resources for foreign tourists as nonresidents, we compared the results of psychological evaluation of tourist spots in Hiroshima obtained by the image-based and onsite experiments using Chinese and Japanese students as nonresidents and residents.

 2. Outline of the experiment In this study, image-based evaluation experiment by Photo Scenes (PS) and onsite evaluation experiment were conducted. In the former experiment, we presented landscape photos of 32 tourist spots in Hiroshima and the explanatory texts about them, and then asked subjects to evaluate each tourist spots. The latter experiment contained two parts: the caption evaluation experiment and the Actual Scene evaluation experiment (AS), which asked subjects to view the same scene as the photo of each tourist spot, and to evaluate in the same way as PS. To figure out the difference between the impression from the photo and the explanatory texts and that from the actual space, we discussed the results of PS and AS out of the three experiments.

 3. Subject recognition As a result of the recognition and experience of the tourist spot, it turned out that there is a big difference in knowledge and experience to the target tourist spot between the Japanese and Chinese subjects. Therefore, the validity of the Chinese and Japanese students as nonresident and resident subjects was confirmed.

 4. Evaluation result of the tourist spots According to the average evaluation score profiles, in case of the Chinese, “willingness to visit” was highly evaluated for famous tourist spots in PS, because they expected more based on a lot of information about the spots beforehand. However, in AS, “satisfaction” was evaluated lower, because they felt the spots were not as good as they had expected, when they actually visited.

 5. Comparison of the psychological evaluation structure The factor analysis was applied to the image evaluation based on the Semantic Differential techniques and extracted six factors such as “Inherency”, “pleasantness”, “Traditionality” “Openness”, “Harmony”, and “Affinity”. According to the result of multiple regression analysis, it can be said that the Chinese didn’t have abundant knowledge about Japan and Japanese tourist spots, so they were interested in the spots with the special characteristics that they perceived simply regardless whether it is Japanese or not Japanese before visiting. However, through the visiting, they recognized more the openness of the spot and understood more the likeness 　of Japan, and as a result, they became to have an interest in Japanese traditional elements.

 6. Conclusion The tendencies of psychological evaluation by Chinese subjects as nonresidents were characterized by comparing with that by Japanese as residents. In the future, it will be necessary to clarify the evaluation structure, such as relevance to the results of caption experiment.

 In this study, we classified 16 kinds of arrows which are frequently found at the station as "straight arrows" and "polygonal arrows". Then we conducted a fundamental experiment using arrows as the sole experiment elements of the space. In this experiment, subjects made a direction judgment and wrote a progress route intuitively. In this way, we investigated the factors of causes that made people get lost at the station. The results we got from the experiment are summarized below.

 1) As for all the 8 types of "straight arrows", subjects are apt to judge the direction as a plan indication rather than an elevation one.

 2) The downward "straight arrows" may cause misunderstandings in the direction judgment.

 3) In case of "straight arrows", except of rightward and leftward arrows, "It is difficult to do the judgment between the plan and the elevation indications." became the most important reason why the subjects got lost in the direction judgment.

 4) Based on the experiment of "polygonal arrows", we found that some upward arrows worked particularly better indicating the direction and progressing the route. Also, subjects can easily understand these signs.

 5) The downward "polygonal arrows" may cause misunderstandings in the direction judgment.

 6) As for all of these 8 types of "polygonal arrows", "It is difficult to know the distance to the turning point." became the most important reason why the subjects get lost in the direction judgment.

 7) In some cases of "polygonal arrows", the users may interpret them as elevation indications which the railway company does not expect.

 8) It is possible that the U-turn arrows of "backward progress" indications are not working properly.

 In both of the "straight arrows" and the "polygonal arrows", the force of indications of the downward arrows is weaker than the upward ones. And the downward arrows are difficult to understand. It is necessary to investigate the causes in detail in the future. In addition, as for the "straight arrows", it was obvious that the users may confuse the direction of plan indications with the elevation ones. This became a factor which made people get lost at the station. According to "The Ecological Approach To Visual Perception", we think that the force of the indication will be increased by incorporating the linear perspective and the texture gradient in the signboard. However, this idea needs to be examined by more experiments in the future. Otherwise, when applying to the real situation of the space, there are more factors other than the arrow signs in the view range of the direction judgment, so we need to consider this influence multifaceted.

 It is vitally important to estimate visibility of all visual targets in real environments to maintain the visual safety. As luminance images or luminous color images in real lit environments are relatively easily obtained, it is reasonable to attempt to establish a method to estimate visibilities of objects using these images.

 Authors have proposed C-A graph that can estimate visibility from luminance images using contrast profile method. In another previous study, authors have proposed a method to analyze effects of color using XYZ stimulus values and L*a*b* color space. The proposed method calculates luminous contrast values (LC, aC, bC) and luminous average values (LA, aA, bA) from XYZ stimulus values by referring to L*a*b* color space. ‘LC’ indicates the luminance contrast, ‘aC’ indicates the red-green contrast and ‘bC’ indicates the yellow-blue contrast.

 In addition, authors have conducted visibility evaluation experiments using circular objects, varying luminance and chromaticity for the common color vision type people. Circular objects and background with varying background luminance, background chromaticity, luminance contrast (LC), chromaticity contrast (aC, bC), and object size were displayed on the 27 inch digital display screen. Subjects observed circular objects displayed on uniform-luminance background with both eyes from 2500 mm point. Subjects evaluated visibilities of circular objects using 5-step scale of ‘0’ to ‘4’. ‘0’ is invisible, ‘1’ is visible with considerable effort, ‘2’ is visible with effort, ‘3’ is visible effortlessly and ‘4’ is easily visible. Experiment results were presented on CIE 1931 xy chromaticity diagram and aC-bC graph.

 In this study, same experiments were conducted for congenital protan (protanopia/protanomaly) and deutan (deuteranopia/deuteranomaly) types in order to achieve color universal design. Experiment results were analyzed using CIE 1931 xy chromaticity diagram and aC-bC graph and experiment results of common, protan, and deutan types were compared respectively. The findings of the study are listed below.

 1) When circular object size is 5 minutes, background chromaticity was D65, and luminance contrast (LC) was zero, tendencies of results of protan and deutan types were similar with those of the common type with slightly greater range in red-green direction. Visibilities of circular objects with the color range from greenish yellow to violet worsened by the effect of small-field tritanopia.

 2) The same tendency was seen in other background chromaticities.

 3) When circular object sizes became greater than 5 minutes, tendencies of results of protan and deutan types had greater range in red-green direction than those of the common type.

 4) When luminance and chromaticity contrasts were changed simultaneously, it was seen that tendencies of results of common, protan, and deutan types differed greatly. As the LC absolute values increased, visibilities of circular objects of the common type improved overall, however, those of protan and deutan types worsened in specific chromaticity contrast (aC, bC) ranges. When LC is positive (LC>0), visibilities of circular objects of the protan type worsened in contrast ranges that are redder than ‘color confusion line of S-cone’, while visibilities of circular objects of the deutan type worsened in contrast ranges that are greener than ‘color confusion line of S-cone’.

 In the next step, we will calculate color contrast values from LMS stimulus values and establish the visibility estimation method considering color vision types.

 According to the Meteorological Agency 'Climate Change Monitoring Report 2016', the average annual temperature in Japan has risen.

 The urban area reports that it is improving the heat island.

 The authors focused on the influence of recent weather changes on cooling and heating load.

 The heating and cooling loads are calculated from weather change and building condition, internal heat generation, air conditioning driving situation.

 The authors created data from JMA observation data of 30 locations nationwide from 1991 to 2017 (27 years) recently for heat load calculation.

 Using these data, the cooling and heating loads of the detached house were strictly calculated using the thermal load calculation program (LESCOM-mint).

 Summary

 ① Annual cooling loads increase both in cities and elsewhere.

 The biggest point of increase is Fukuoka's year 0.838 GJ, followed by Nagasaki, Ishigaki-jima, Cyoshi, Tokyo.

 ② Annual heating loads are decreasing due to secular change, but there are some points where it is increasing.

 This is because the outside air temperature that greatly affects the heating load does not necessarily increase due to secular change.

 The major reduction point is Sendai's year 0.102 GJ, followed by Wakkanai, Abashiri, Kagoshima, Morioka.

 The biggest point of increase is 0.0316 GJ in Nagasaki year, then Sapporo, Tokyo, Niigata and Nagoya.

 The rate of increase is small compared to the rate of decrease.

 ③ The instantaneous cooling loads increase at 29 locations, excluding one point Kofu.

 The biggest point of increase is Hakodate year 0.0338 kW, followed by Morioka, Ishigaki-jima, Aomori, Akita in order.

 Kofu is decreasing, but it is very small with a decrease of 0.0061 kW a year.

 ④ The instantaneous heating loads are increasing and decreasing over time.

 The increase is due to the fact that the fluctuation of the outside air temperature in the extremely cold season in winter is different from year to year, and it does not rise monotonically with aging.

 The major point of decrease is 0.0663 kW of Ishigaki-jima, then Miyakojima, Aomori, Naze, Abashiri in order.

 The biggest point of increase is Nagasaki's year 0.0632 kW, followed by Hiroshima, Matsuyama, Kumamoto and Osaka.

 In this research, a simple heat load calculation method that employs phase changing material (PCM) replacement model for outer or inner thermal storage wall, ceiling, and floor is proposed.

 Chapter 1 describes the purpose of this research. It is proposing a simplified method to calculate room temperatures and heat load for a model house that contains PCM in its outer or inner thermal storage wall, ceiling or floor.

 The proposed simple calculation method is suggested in Chapter 2. Regarding the replacement of PCM with internal thermal mathematical models for adjoining rooms, we note that the replaced PCM temperature and room temperature affect each other. The proposed model assumes that PCM absorbs solar radiation heat.

 In Chapters 3 and 4, the calculation results for 1,728 replacement model cases were verified and compared to a detailed calculation method, and the accuracy of the simplified method was confirmed. Therefore, it was confirmed that, for cases where the PCM phase change temperature was higher than the set heating temperature, room temperature results with the replaced model demonstrate a tendency similar to that of the detailed calculation. Differences in room temperature results were because the timing of the end of the phase change sometimes differed between the detailed model and the replacement model. However, the phase change timing differences were not significant, and its effect on the annual room temperature was limited. The detailed calculation and replacement model results in the heating and cooling load also indicated similar trends. Conversely, cases where the PCM phase change temperature is not higher than the set heating temperature, differences between the detailed calculation and the replacement model for room temperature and heat load were significant because the replacement model assumes that PCM absorbs solar radiation heat rather than heat from the air in the room.

 As discussed in Chapter 5, there are some issues involved in applying this replacement model to the thermal load calculation of a house.

 Introduction

 The number of residences with thick insulation and air-tightness continues to increase in cold region. Ventilation loads cause over half of the heating load of such houses. In the previous report¹⁾, we developed a ventilation preheating system combining the solar thermal collector and the PCM panel shown in Fig. 1. We verified its performance with the experiment and numerical simulation. As the result, it is difficult to optimize the control system by a simple feedback system against ever-changing weather and indoor and outdoor environments. Therefore, in this research, we aim to develop a control method for a fan air volume of ventilation preheating system by machine learning. Initially, we outlined machine learning and reinforcement learning. Next, we explained how to introduce reinforcement learning in existing systems. Finally, we examined the performance of existing ventilation preheating system (VP system) and ventilation preheating system (RL system) controlled by reinforcement learning and examined the possibility of practical application.

 Introduction of Reinforcement learning

 Fig. 2 and Fig. 3 shows the outline of reinforcement learning. We used the Q-Learning algorithm as the method of reinforcement learning. Fig. 4 shows the calculation process and simplified code of Q-Learning. The calculation formulas and algorithms were used to install reinforcement learning into the existing system. Fig. 6 shows a flowchart of the Q-Learning algorithm used in the calculation process in this study.

 Control of ventilation preheating system with reinforcement learning

 We set the target schedule from October to March and prepared several cases of the operating periods and the air volume in VP system. We compared those results with the results operated by RL system. The findings obtained by this study are shown below.

 (1) As shown in Fig. 7 and 8, it was confirmed that the RL system automatically controlled the air flow of the fan.

 (2) As shown in Fig. 9, the RL system is equal to or higher than the performance of the VP system through the calculation period. Especially in the winter season, December to February, the performance of the RL system was the best in all cases.

 (3) Since reinforcement learning was successfully introduced. Also, the performance of RL system was better than the other cases, it is possible to study for practical application in the future.

 Urban heat islands (UHIs) represent a critical thermal environmental problem in cities. In Japan, UHIs have been mostly focused as an issue in large cities. However, it is currently considered a serious problem in medium-sized cities as well. The outdoor thermal conditions in urban areas of these cities are also declining. To counteract the UHI effect, it is necessary to understand the conditions of outdoor environments in areas of concern and determine a solution specific to each area. Sensible heat load and outdoor thermal comfort are evaluated, via the heat island potential (an index based on the sensible heat flux from all external surfaces of an area to the atmosphere) and the mean radiant temperature (considering solar radiation), respectively. Herein, we discuss one of the medium-sized cities, Takamatsu, located on the coast of the Seto Inland Sea. In recent years, Takamatsu has been undergoing urban redevelopments in a central urban area. At the same time, a land readjustment project has been implemented in a suburban area.

 This study demonstrates the role of spatial components in the UHI effect and examines solutions to prevent the subsequent degradation of the outdoor thermal environment in medium-sized cities. Three-dimensional models replicating typical spaces of Takamatsu were constructed with reference to districts in the city. Heat balance simulations using these models revealed the conditions of the outdoor environment in Takamatsu. The following issues were identified: 1) There are many asphalt-paved streets and parking lots, where the surface temperature increases under sun exposure. 2) The water surface area is small. 3) There are few trees and hence lesser cooling effect. 4) There is little consideration for thermal performance of building finishes.

 Proposals to solve these issues (related to land cover, trees, and building finishes, which are major spatial components) are consistent with the climate, culture, and history of the city. Changes in the outdoor thermal environment by implementing the proposals revealed the significance of spatial components in determining the outdoor thermal conditions.

 Based on these results, the following observations were made: 1) In the southern part of the urban area in Takamatsu, the neighboring districts, which are primarily residential areas, have different spatial components and outdoor thermal environments. There are diverse outdoor spaces with different characteristics in emerging city areas. 2) In the same area, asphalt-paved streets and parking lots account for more than 30% of the district, even in areas where parking lots attached to large stores do not exist. Conversely, there are few trees, and they represent approximately 5% or less of the district, by using the horizontally projected area of the crowns of medium-height and tall trees as an index. 3) By implementing various proposals related to spatial components, the maximum value of the heat island potential (at 1:00 p.m.) decreased by 7.7 °C in District 1 and 7.2 °C in District 2. Areas where the mean radiant temperature decreased were spread throughout. In some shaded areas, the mean radiant temperature further decreased to approximately 40 °C, reducing the difference from the air temperature (36.9 °C at 1:00 p.m.) to approximately 3 °C.

 By understanding the characteristics of the area and implementing dedicated approaches to address the problems, it is possible to reform the outdoor thermal environment in urban areas and revolutionize the lifestyle in cities.

 In recent years, the summer thermal environment in urbanized areas is becoming severe inhospitable for residents because of urban heat island effect and global warming. These phenomena are collectively regarded as urban warming. Many countermeasures against urban warming and its effects have been studied. However, these countermeasures should be installed in suitable places for effective urban environmental planning.

 From this background, “urban environmental climate maps” shows climatic characteristics and countermeasures for each zone are proposed as effective tools to inform urban planning stakeholders about the urban thermal environment. As one layer of urban environmental climate maps, zoning maps classifying urban areas in terms of urban climate are necessary for presenting and considering effective countermeasures for each zone.

 Using results of numerical calculation by meso-scale meteorological models is apparently effective to produce climatic zoning maps. However, spatial resolution of such numerical calculation is relatively rough, and the influence of microtopography can’t be reflected. In some cases, observation data is utilized for classifying target area. But in this case, it is also difficult to reflect complicated topography because of limitations on the number of measurement points.

 As described above, the spatial resolution is rough in numerical calculations, and the observation has limitations on the number of observation points, so it is considered that both are insufficient to consider the influence of microtopography. However, in hilly cities like Yokohama, there are many small valleys and the complicate topography, so it seems that there are some effects by microtopography on urban climate. Therefore, it is considered that more detailed understanding and zoning are necessary.

 In this study, the goal is to produce an urban environmental climate map that considers microtopography effects in Yokohama. Firstly, we calculate the urban climate in a wider area and high resolution by using the Multi-Scale Simulator for the Geoenvironment installed in the Earth Simulator, to know the current situation of the urban climate in Yokohama. Furthermore, principal component analysis and cluster analysis with the calculation result as input were carried out, and climatic zoning map including microtopography effect in Yokohama was produced. Authors also analyzed the climatic characteristics of each zone and confirmed the accuracy of zoning through comparison to the observation results.

 The major findings are as follows:

 1) In addition to the air temperature reduction effect by sea breeze, the air temperature tends to be higher in the valley than in the surrounding area. The air temperature is also higher in the valley only in inland at 13 o'clock. It seems that this tendency is more conspicuous in weak wind environments.

 2) Three principal components were extracted by principal component analysis. These principal components related to the sea breeze from Tokyo Bay and nighttime sea water temperature, the sea breeze from Sagami Bay and the microtopography respectively.

 3) Urban meshes in Yokohama are classified into five zones by cluster analysis and authors produced climatic zoning map of Yokohama. The zones were mainly classified by the distance from the Sagami Bay / Tokyo Bay and the microtopography. The overall patterns were mostly same.

 Computer Graphics are widely used as a powerful tool for design study at architectural design field. To visualize planning buildings by Computer Graphics at designing stage before its construction helps to enhance design quality and shorten designing term. On the other hand, Computer Graphics technologies become increasingly sophisticated and computation time to generate images goes on increasing accordingly. It is not uncommon to take tens of minutes to generate one image. Increase of rendering time prevents frequent use of Computer Graphics. Instant image generation is necessary to utilize Computer Graphics at design stage and dramatic improvement of rendering speed is highly required by architectural designer.

 Calculation speed capability of CPU, that is heart of personal computer, currently faces limit through previous drastic progress and major progress of CPU speed can no longer be expected. In such condition of computer GPU (Graphic Processing Unit) recently attracts attention. GPU is less flexible than CPU on application but has thousands of cores for parallel computing. If we can take advantage of the characteristics of GPU, there is the big potential of high performance calculation to overtake CPU capability.

 We pursue developing a practical super high speed Computer Graphics rendering system whose calculation speed is more than hundreds times faster than ordinal rendering systems by CPU calculation, and going to provide the system through internet for designers to use high end Computer Graphics images quickly and easily at design study. This paper proposes an efficient parallel calculation method to extract high speed characteristics of GPU, and judges effect of the proposal method by implementing the method to rendering software and timing its calculation speed to generate images to confirm its practicality.

 GPU, namely parallel processing, has advantage at same calculation huge times, but has weak point at branched operation. It is basically difficult for Computer Graphics rendering software, that is constructed by huge branched operations for various rendering functions, to get benefit of high speed performance by GPU. Therefore we propose a hybrid rendering method of CPU calculation and GPU calculation. We extract small process of heavy load calculation from total rendering process, and apply GPU calculation to the heavy load process only. By calculating the rest using CPU huge branched operations by GPU is avoided and we can get benefit of GPU high speed calculation.

 Furthermore, for efficient calculation using multi GPUs we propose equal load distribution method to get rid of variation of calculation load at each GPU and finish calculation simultaneously at all GPUs.

 We implemented these proposed method for GPU calculation to CPU version of global illumination rendering software based on path tracing method. We timed rendering calculation of several test images by the GPU version of the software with a server system consisted of 7 GPUs, and got result of faster hundreds times to thousands times than CPU version of the same software.