Odour emissions from complex industrial plants may cause potential impacts on the surrounding areas. Consequently, the validation of effective tools for the control of the associated environmental pressures, without hindering economic growth, is strongly needed. Nowadays, senso-instrumental methods by using Instrumental Odour Emissions Systems (IOMSs) is among the most attractive tool for the continuous monitoring of environmental odours, allowing the possibility of obtaining real-time information to support the decision-making process and proactive approach. The systems complexity and scarcity of real data limited their wider full-scale employment. The study presents an advanced prototype of IOMS for the continuous classification and quantification of the odours emitted in ambient air by complex industrial plants, to continuously control the plants emissions with backwards approach. The IOMS device was designed and optimized and included the system for the automatic control of the conditions inside the measurement chamber. The designed operational procedures were presented and discussed. Results highlighted the influence of temperature and air flow rate for the measurement repeatability. Accurate prediction model was created and optimized and resulted able to distinguish 3 different industrial odour sources with accuracy approximately equal to 96%. The models were optimized thanks to the software features, which allowed to automatically apply the designed statistical procedures on the identified dataset with different pre-processing approach. The usefulness of having a fully-developed and user-friendly flexible system that allowed to select and automatically compare different settings options, including the different feature extraction methods, was demonstrated in order to identify the best prediction model.

复杂工厂的气味排放可能会对周围地区造成潜在影响。因此，强烈需要在不妨碍经济增长的情况下验证用于控制相关环境压力的有效工具。如今，使用仪器气味排放系统（IOMS）的传感仪器方法是对环境气味进行连续监测的最有吸引力的工具之一，从而可以获取实时信息以支持决策过程和主动方法。系统的复杂性和实际数据的稀缺性限制了它们更广泛的全面使用。这项研究提出了一种IOMS的高级原型，用于对复杂的工厂，工厂，工厂，环境空气中散发的气味进行连续分类和量化，以向后的方法连续控制工厂的排放。IOMS设备经过了设计和优化，包括用于自动控制测量室内条件的系统。提出并讨论了设计的操作程序。结果突出了温度和空气流速对测量重复性的影响。创建并优化了准确的预测模型，结果能够区分3种不同的工业气味源，准确度约等于96％。通过软件功能对模型进行了优化，从而可以使用不同的预处理方法在识别出的数据集上自动应用设计的统计程序。