Energy Consumption Analysis and Modeling of Cannabis Cultivation Facilities**

自2018年大麻合法化以来，由于需要加强过程控制和产量、病虫害管理，室内大麻种植的能源需求加速增长。尽管有关加拿大大麻生长所需能源需求的数据有限，但美国的研究发现，大麻生产占全国用电量的1%，即每年60亿美元。因此，通过实时监控设备跟踪大麻种植设施的能源消耗对于了解设施的操作实践与能源消耗行为之间的相关性至关重要。行业合作伙伴bringrid提供无线分析系统，为精准农业提供实时洞察、警报和报告。该系统利用传感器来测量生长设施内的变量。这些可能包括能源使用、空气温度、湿度和二氧化碳水平、土壤或养分浓度和pH值等等。bringrid正在与约克大学的SGRL团队合作开展该项目，开发一种流程，该流程可以通过传感器接收从耕种器收集的实时能量数据，并执行负载分解，以区分单个设备类别对整体能源使用的贡献。此外，还将开发一个数学模型，建议对操作实践进行更改，从而最终使能源消耗最小化。该模型将被制定为一个优化问题，其中将考虑几个目标函数：净利润最大化，大麻产量最大化，能源消耗最小化，温室气体排放最小化。建议的模式会考虑各种运作限制，以满足设施的运作要求，例如不断增长的需求和产品质量。将使用从安大略省现有大麻设施收集的实时数据，对不同的案例研究测试拟议模型的有效性

Since legalization in 2018, indoor Cannabis cultivation has seen an accelerated growth in energy demand due to the need for greater process control and yield, pest, and disease management. Although data about the required energy demand for cannabis growth in Canada is limited, studies in US found that cannabis production is responsible for 1% of national electricity use, or USD $ 6 billion each year. Hence, tracking the energy consumption for cannabis cultivation facilities via real-time monitoring devices is paramount for understanding the correlation between the operation practises of the facilities and the energy consumption behaviour. The industry partner, Braingrid, provides a wireless analytics system that delivers real-time insights, alarms and reports for precision agriculture. The system leverages sensors to measure variables within the grow facility. These can include energy use, air temperature, humidity and carbon dioxide levels, soil or nutrient concentration and pH, and more. Braingrid is working on this project with the SGRL team at York University to develop a process that can receive real-time energy data collected from the cultivator facilitates via sensors and perform load disaggregation to distinguish the contribution of individual equipment categories to the overall energy use. Also, a mathematical model that suggests changes to the operation practises, which will eventually lead to the minimization of the energy consumption, will be developed. The model will be formulated as an optimization problem, where several objective functions will be considered: maximize the net profit, maximize the cannabis yield, minimize the energy consumption, and minimize the greenhouse gas emissions. The proposed model would take into account various operating constraints to satisfy the operation requirements of the facility e.g., growing needs and product quality. The efficacy of the proposed model will be tested for different case studies using real-time data collected from existing cannabis facilities in Ontario.**